-
Notifications
You must be signed in to change notification settings - Fork 1
Expand file tree
/
Copy pathschedule.html
More file actions
785 lines (753 loc) · 74.7 KB
/
schedule.html
File metadata and controls
785 lines (753 loc) · 74.7 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
<!DOCTYPE HTML>
<!--
Spatial by TEMPLATED
templated.co @templatedco
Released for free under the Creative Commons Attribution 3.0 license (templated.co/license)
-->
<html>
<head>
<title>Schedule - ML4PSP</title>
<meta charset="utf-8" />
<meta name="viewport" content="width=device-width, initial-scale=1" />
<link rel="stylesheet" href="assets/css/main.css" />
</head>
<body>
<!-- Header -->
<header id="header">
<h1><strong><a href="index.html">ML4PSP</a></strong></h1>
<nav id="nav">
<ul>
<li><a href="index.html">Home</a></li>
<li><a href="schedule.html">Schedule</a></li>
<li><a href="join.html">Join</a></li>
</ul>
</nav>
</header>
<a href="#menu" class="navPanelToggle"><span class="fa fa-bars"></span></a>
<!-- Main -->
<section id="main" class="wrapper">
<div class="container">
<header class="major special">
<h2>Upcoming Seminar</h2>
</header>
</section>
<!-- Schedule -->
<section id="sched" class="wrapper style2 special">
<h3> 2026 Seminar Series </h3>
<h3>April 21, 2026</h3>
<p> Tuesday at 9 AM US Pacific </p>
<div class="container">
<!-- <div class="6u 12u$(xsmall)"> -->
<div class="image fit captioned">
<h3> Jichao Feng</br>Northern Illinois University</h3>
<p>Planetary-Scale Similarity Search for Mars Orbital Imagery with Foundation-Model Embeddings</p>
<p>Mars orbital archives now contain enough imagery that finding morphologically similar features is bottlenecked by search, not data. We present a planetary-scale similarity search system built on foundation-model embeddings over the full CTX Murray global mosaic (~26.9M indexed locations). A Vision Transformer pretrained via self-supervised learning on millions of CTX patches produces embeddings that capture surface texture and landform semantics without any labels. Deployed as a quantized vector index on a single server, the system supports sub-second instance-level retrieval ("find terrains like this"), geo-filtered search within regions of interest, and interactive relevance feedback for iterative refinement. The system is publicly accessible at findmars.space.</p>
</div>
</div>
<!-- <p> Please check back here for updates or join our <a href="join.html"> listserv</a>. <p> -->
<!-- Comment for new organizers, below is where you will update with details of new speakers -->
<p>Previous seminars with <a href="schedule.html">abstracts</a> can be found on <a href="https://studio.youtube.com/channel/UCyS8UU1a6K0NM_lSN7ykAnA/playlists">YouTube</a>. </p>
</section>
<!-- Past -->
<section id="main" class="wrapper">
<div class="container">
<header class="major special">
<h2>Previous Seminars</h2>
</header>
</section>
<section id="sched" class="wrapper style2 special">
<!-- Previous Talks -->
<!-- Comment for new organizers, below is where you will move previous seminars to when they are done, I've added as a comment a
new section for you to fill in when it's time -->
<details>
<summary> <font size="+1"> <b> SEMINAR SERIES 2024 </b> </font>
<p> <a href="https://www.youtube.com/playlist?list=PL4kOs9mVzcpa1Jo1oqPSIdo_woKyc5-6w"> Recordings </a> </p>
</summary>
<details>
<summary><font size="+1">June 4, 2024 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Michael Holland </br> Brigham Young University </h3>
<p> This Crater Does Not Exist: How Synthetically Created Craters Can Help Us Understand Crater Formation and Key Crater Features. </p>
<p> In the study of impact cratering machine learning has historically been used for crater segmentation and the discovery of new craters. With the advent of new and improved techniques, deep learning can also be used to enhance the current understanding of crater geomorphology. State-of-the-art machine vision techniques allow scientists to process and understand satellite imagery on scales previously thought impossible. With the use of unsupervised generative algorithms such as Diffusion Models [1], Generative Adversarial Networks (GANS) [2], and other contemporary generative models, higher-order features of impact cratering can be learned and explicitly modeled. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Andreas Bechtold </br> Austrian Academy of Sciences and University of Vienna </h3>
<p> A simulation-based deep learning training approach for autonomous detection of shatter cones in Mars rover images. </p>
<p> Shatter cones are rocks with characteristic features that can form during high-velocity impact events of asteroids or comets and are the only known macroscopic indicator for shock metamorphism. They have been involved in the identification of impact craters on Earth, and were noted in a few meteorites. Although the Moon and Mars are largely covered with impact structures, shatter cones have never been unambiguously identified during any space missions to these planetary bodies. The specific morphological features that characterize shatter cones may represent an interesting criterion in terms of image recognition. In a feasibility study we, therefore, explored how deep learning algorithms can be trained towards an automatic detection of shatter cones in images from Mars rover cameras using a simulation-based training approach. Deep learning requires large amounts of training data. As images of shatter cones from the surface of Mars do not exist, we have virtually placed shatter cones in a real contextual environment to produce the necessary training images. The feasibility study has shown that the approach with shatter cones in artificial Mars rover scenes is in principle suitable to train neural networks for automatic detection of shatter cones and provided indications of what can be done to improve the results. </p>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> April 9, 2024 </font>
</summary>
<div class="container">
<div class="row 300%">
<!-- <div class="6u 12u$(xsmall)"> -->
<div class="image fit captioned">
<h3> Opal Issan </br> University of California San Diego </h3>
<p>Bayesian Inference and Global Sensitivity Analysis for Ambient Solar Wind Prediction</p>
<p>The ambient solar wind plays a significant role in propagating interplanetary coronal mass ejections and is an important driver of space weather geomagnetic storms. A computationally efficient and widely used method to predict the ambient solar wind radial velocity near Earth involves coupling three models: Potential Field Source Surface, Wang-Sheeley-Arge (WSA), and Heliospheric Upwind eXtrapolation. However, the model chain has 11 uncertain parameters that are mainly non-physical due to empirical relations and simplified physics assumptions. Therefore, we propose a comprehensive uncertainty quantification (UQ) framework that can successfully quantify and reduce parametric uncertainties in the model chain. The UQ framework utilizes variance-based global sensitivity analysis followed by Bayesian inference via Markov chain Monte Carlo to learn the posterior densities of the most influential parameters. The sensitivity analysis results indicate that the five most influential parameters are all WSA parameters. Additionally, we show that the posterior densities of such influential parameters vary greatly from one Carrington rotation to the next. The influential parameters are trying to overcompensate for the missing physics in the model chain, highlighting the need to enhance the robustness of the model chain to the choice of WSA parameters. The ensemble predictions generated from the learned posterior densities significantly reduce the uncertainty in solar wind velocity predictions near Earth.</p>
</div>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
<details>
<summary> <font size="+1"> February 6, 2024 </font>
</summary>
<div class="container">
<div class="row 300%">
<!-- <div class="6u 12u$(xsmall)"> -->
<div class="image fit captioned">
<h3> Dattaraj Dhuri </br> New York University Abu Dhabi </h3>
<p>Characterizing proton auroras on Mars using explainable Machine Learning</p>
<p>Proton auroras occur frequently on the dayside of Mars. These are seen as an intensity enhancement of Ly‐alpha (121.6 nm) emission between ~ 110 – 150~km altitudes. These auroras are triggered primarily by electron stripping and charge exchange between solar wind protons and the neutral hydrogen in the corona; forming the energetic neutral hydrogen atoms (ENA) that penetrate down to the thermosphere. Once in the thermosphere, these ENAs de-excite by collisions with the atmosphere and release the aurora emission. Recent observations of “patchy” proton auroras suggest additional mechanisms could be at play, possibly depositing solar wind protons directly in Mars’ thermosphere to affect these spatially localized proton auroras. In this talk, I will discuss our work on developing a data-driven, artificial neural network (ANN) model to characterize proton aurora occurrences. The ANN is developed using Mars Atmosphere and Volatile EvolutioN (MAVEN) in-situ observations and limb scans of Ly-alpha emissions between 2014 -- 2022. We show that the ANN reproduces individual Ly-alpha intensities with a Pearson correlation of ~ 95% along with a faithful reconstruction of the observed Ly-alpha emission altitude profiles. We use SHapley Additive exPlanations (SHAP) analysis to discover relationships between the inputs and the modeled Ly-alpha intensities. We find that Solar Zenith Angle, seasonal CO2 atmosphere variability, solar wind temperature, and density are the most important features for the modeled proton auroras. We also demonstrate that such a data-driven model can serve as an inexpensive tool for simulating and characterizing Ly-alpha response under a variety of seasonal and upstream solar wind conditions.</p>
</div>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
</details>
<details>
<summary> <font size="+1"> <b> SEMINAR SERIES 2022 - 2023 </b> </font>
<p> <a href="https://www.youtube.com/playlist?list=PL4kOs9mVzcpYGD-1ToO5vjzww4E_Ac8kF"> Recordings </a> </p>
</summary>
<details>
<summary> <font size="+1"> October 3, 2023 </font>
</summary>
<div class="container">
<div class="row 300%">
<!-- <div class="6u 12u$(xsmall)"> -->
<div class="image fit captioned">
<h3> Serina Diniega & Umaa Rebbapragada </br> JPL, NASA </h3>
<p>Creating a Global Map of Martian Frost, Using Visible Images and Thermal Data</p>
<p>The winter-time formation, evolution, and retreat of surface frost provide the best global information about the carbon dioxide and water cycles on present-day Mars. Such frost has previously been mapped using coarse-resolution thermal data, but these maps generally miss small-scale patches of frost in the mid- and lower-latitudes. High resolution imagery and spectral frost detections have been studied within a few individual sites, but manual inspection of such datasets over the full globe is too labor-intensive. In our study, we utilize data science techniques to detect frost within all these datasets so as to globally map present-day Martian seasonal frost cycle, with a high-number of ties to landform-scale (10s-meter-scale) environments. This presentation will focus on our work with visible datasets (i.e., >20k HiRISE images and >30k CTX images), where we detect frost via convolutional neural networks (CNNs) with pre-labeled datasets. Furthermore, we will touch on how we use statistical methods to combine the visible frost map with other datasets, decreasing the ambiguity of visible detections and bridging between the different resolutions.</p>
</div>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> September 12, 2023 </font>
</summary>
<div class="container">
<div class="row 300%">
<!-- <div class="6u 12u$(xsmall)"> -->
<div class="image fit captioned">
<h3> Mireia Leon-Dasi </br> Observatoire de Paris, LESIA </h3>
<p>Using deep learning to characterise explosive volcanism on Mercury </p>
<p>The existence of explosive volcanic eruptions on Mercury has been evidenced by the MESSENGER mission through the detection of volcanic vents and associated pyroclastic deposits. However, the diversity of these features in terms of morphology, spectral properties and location, complicates the global characterisation and timing of the volcanic eruptions. In this talk, we introduce the application of an unsupervised deep learning technique (the 3D convolutional autoencoder) to capture the signature of these deposits in the spectral and spatial domains. With this technique we define the extent of the deposits, overcoming the limitations of the traditional visual interpretation approach, and we illustrate how this highlights the differences within and between deposits. Moreover, we dive into the latent space of the architecture to understand the information extracted, using deep learning as an exploratory technique. </p>
</div>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> June 13, 2023 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Abby Azari </br> University of California, Berkeley </h3>
<p> A Virtual Solar Wind Monitor for Mars Using Gaussian Processes </p>
<p> As a spacecraft orbits around a planet it travels through multiple regions and it is often a long time between measurements of any particular region. This is a problem if we wish to understand how the physical processes occurring in one region might affect a different region as we never measure two regions at the same time. I address this problem for understanding how the solar wind affects Mars' space environment by developing a virtual solar wind monitor for Mars using Gaussian processes with MAVEN spacecraft data. In this presentation I discuss how use of this method allows for uncertainty quantification of the solar wind's influence on physical processes in Mars' space environment and discuss potential future implementation at other planets </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Alex Barrett & Jack Wright </br> The Open University & The European Space Agency </h3>
<p> NOAH-H: Characterizing Mars rover landing sites using deep learning </p>
<p> The Novelty or Anomaly Hunter – HiRISE (NOAH-H) is a deep learning convolutional neural network intended to classify high-resolution satellite images of Mars. It was trained to classify terrain at the two finalist European Space Agency ExoMars Rosalind Franklin Rover mission landing sites, Oxia Planum (final selection) and Mawrth Vallis (runner-up). In this talk, a sequel to one given in this seminar series in November 2021, we will introduce NOAH-H and give an update on the latest results from the project. We will show: the NOAH-H Oxia Planum map published in 2022; the transfer of NOAH-H to Jezero crater, our qualitative comparison between NOAH-H and Perseverance rover and Ingenuity helicopter images, our quantitative comparison with a human-made map, and; the NOAH-H map of the Mawrth Vallis landing site, which we have recently submitted for publication. </p>
</div>
</div>
</div>
</div>
<!-- </details> -->
</details>
<!-- <font size="+1"> - </font> -->
<!-- </details> -->
<details>
<summary> <font size="+1"> March 7, 2023 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Laura Breitenfeld </br> Stony Brook University </h3>
<p>Quantifying the Mineralogy of the Asteroid Bennu Using Thermal Emission Spectra and Machine Learning Multivariate Model Applications</p>
<p>Bennu, the target of the OSIRIS-REx mission, is an asteroid with compositions analogous to low petrologic type CI, CM, CR, and/or ungrouped carbonaceous chondrites. Asteroids like Bennu provide information about the building blocks of the early Solar System. Through the collection of a spectral training set and machine learning multivariate analysis models (partial least squares), this work analyzes the remote sensing thermal emission spectral data from the OSIRIS-REx mission to inform mineral quantification on Bennu.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Philippe Garnier </br> University Toulouse </h3>
<p> Ranking the Drivers of the Martian Bow Shock Location </p>
<p>The Martian interaction with the solar wind leads to the formation of a bow shock upstream of the planet. The shock dynamics appears complex, due to the combined influence of external and internal drivers. In this talk, we will compare the influence of the main drivers of the Martian shock location, based on several methods and published datasets from Mars Express (MEX) and Mars Atmosphere Volatile EvolutioN (MAVEN) missions. We include here the influence of the crustal fields, extreme ultraviolet fluxes, solar wind dynamic pressure, magnetosonic Mach number and Interplanetary Magnetic Field parameters. The bias due to the cross-correlations among the possible drivers is investigated with a partial correlations analysis. Several model selection methods (Akaike Information Criterion and Least Absolute Shrinkage Selection Operator regression) are also used to rank the relative importance of the physical parameters. We will also mention several recent works including machine learning, such as the martien shock detection or the identification of magnetic reconnection Electron Diffusion Regions at the Earths magnetopause.</p>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> May 2, 2023 </font>
</summary>
<div class="container">
<div class="row 300%">
<!-- <div class="6u 12u$(xsmall)"> -->
<div class="image fit captioned">
<h3> Cai Ytsma </br> Cai Consulting / University College London </h3>
<p>Fundamental analyses of quantification accuracy in rock spectra using machine learning</p>
<p>With a focus on laser-induced breakdown spectroscopy (LIBS) applications on Mars with ChemCam, I have studied various influences on elemental quantification in rocks using a suite of LIBS spectra from almost 3,000 geological standards. Rock spectra are notoriously complex and it is difficult to build robust and accurate predictive models for unknowns on remote surfaces. From this work I have discovered pros and cons of various machine learning methods, effects of data set size and diversity, and compared results to traditional methods. The moral of the story is: the more data, the better!</p>
</div>
</div>
<!-- <div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Philippe Garnier </br> University Toulouse </h3>
<p> Ranking the Drivers of the Martian Bow Shock Location </p>
<p>The Martian interaction with the solar wind leads to the formation of a bow shock upstream of the planet. The shock dynamics appears complex, due to the combined influence of external and internal drivers. In this talk, we will compare the influence of the main drivers of the Martian shock location, based on several methods and published datasets from Mars Express (MEX) and Mars Atmosphere Volatile EvolutioN (MAVEN) missions. We include here the influence of the crustal fields, extreme ultraviolet fluxes, solar wind dynamic pressure, magnetosonic Mach number and Interplanetary Magnetic Field parameters. The bias due to the cross-correlations among the possible drivers is investigated with a partial correlations analysis. Several model selection methods (Akaike Information Criterion and Least Absolute Shrinkage Selection Operator regression) are also used to rank the relative importance of the physical parameters. We will also mention several recent works including machine learning, such as the martien shock detection or the identification of magnetic reconnection Electron Diffusion Regions at the Earths magnetopause.</p>
</div> -->
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
</details>
<details>
<summary> <font size="+1"> <b> SEMINAR SERIES 2021 - 2022 </b> </font>
<p> <a href="https://www.youtube.com/playlist?list=PL4kOs9mVzcpYiJ1YNd4g5HIdsj3FOQ0Rm"> Recordings </a> </p>
</summary>
<details>
<summary> <font size="+1"> September 28, 2021 </font>
</summary>
<p>Special Seminar - Cross-Listed with the <a href="https://www.openplanetary.org/">OpenPlanetary</a> Seminar Series </p>
<div class="container">
<div class="row 300%">
<!--<div class="6u 12u$(xsmall)">commented out to spread over row-->
<div class="image fit captioned">
<h3> Annie Didier </br> NASA Jet Propulsion Laboratory </h3>
<p> Incepting Interplanetary “Google Search” through Machine Learning </p>
<p>Spacecraft can produce a far greater volume of data than can be downlinked.
Though interplanetary communication rates have grown in orders of magnitude since early missions, they are far surpassed by the growth in data volume produced by on-board instruments.
To bypass the communication bottleneck between spacecraft and ground while optimizing scientific yield, we propose the concept of ‘Interplanetary Google Search,’ a novel approach to spacecraft data retrieval inspired by the Google search engine.
We envision a selective downlink capability with on-board indexing and search where scientists can query a spacecraft’s on-board database for specific, relevant information.
To realize this on-board data storage and indexing vision, we must first introduce the means to extract features relevant to scientific interest from historic data payloads.
The key to our approach is utilizing machine learning to extract features and summarize data.
We have demonstrated this capability using image segmentation and image captioning models of MSL RGB imagery.
Such methods would, for instance, enable scientists to download a full textual summary of the imagery taken by a rover and use this information to downlink data with specific features for further analysis.
This concept has even more potential for a wider range of deep-space missions with more data-intensive instruments (e.g. ground-penetrating radar and hyperspectral imagers), and can be realized with the application of data science.</p>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> October 26, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Mayur Bakrania </br> University College London </h3>
<p>Applying unsupervised learning and outlier detection methods to characterise magnetotail electrons</p>
<p>Collisionless space plasma environments are characterised by distinct particle populations that typically do not mix.
Although moments of their velocity distributions help in distinguishing different plasma regimes, the distribution functions themselves provide more comprehensive information about the plasma state.
Unlike moments, however, distributions are not easily characterised by a small number of parameters, making their classification more difficult.
To perform this classification, we distinguish between the different plasma regions by applying dimensionality reduction and clustering methods to electron distributions in pitch angle and energy space.
The automated classification of different regions in space plasma environments provides a useful tool to identify the physical processes governing particle populations in near-Earth space.
Using outlier detection methods, we can identify anomalous distributions in the magnetotail that are consistent with simulations of the tearing instability.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Jay Laura </br> U.S. Geological Survey </h3>
<p>Planetary Spatial Data Infrastructure and Analysis Ready Data </br> - </p>
<p>Spatial data efficacy is a primary concern for any scientists making use of spatial data products.
The field of planetary spatial data infrastructure (PSDI) includes work to classify and identify foundational data products with well communicated spatial accuracies to ensure appropriate data use.
Under the auspices of PSDI, the nascent push for analysis ready data (ARD) provides a rich opportunity space where data sets are both made available in machine learning ready formats and the ML community is actively engaged to improve the discoverability of said data.
I will present a brief overview of PSDI and planetary ARD as they relate to potential ML activities.</p>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> November 23, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Matthew Cheng </br> University College London </h3>
<p> Automated bow shock and magnetopause boundary detection with Cassini </p>
<p> The Cassini mission spent 13 years in orbit around Saturn collecting a wealth of data about its magnetic field and plasma populations.
A catalogue of bow shock and magnetopause boundary crossings is needed to study the structure of Saturn’s magnetosphere and fundamental plasma processes like instabilities.
However, a very challenging aspect is the manual identification of thousands of crossings. It is both time-consuming and prone to human error.
This calls for ways to standardize the detection of these boundaries through automation.
A range of techniques is explored from traditional time-series analysis on the magnetic field and plasma moments data to modern machine learning techniques like evidential deep learning which quantify classification uncertainty applied to electron energy spectrograms.
</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Alexander Barrett </br> The Open University </h3>
<p> Characterizing the ExoMars landing site at Oxia Planum using deep learning </p>
<p> In this study a Deep Learning Convolutional Neural Network was trained to characterize the landscape of Oxia Planum, Mars.
This work was conducted as part of the preparation for the European Space Agency’s upcoming ExoMars Rosalind Franklin Rover mission.
The aim was to develop an ontology of terrain classes, which could be applied across the landing site and beyond, to form a component to traversability analysis, when combined with engineering information about the rover, and with “ground truth” information about how the classes look once Rosalind Franklin lands.
Studies combining high resolution images with large spatial extents, such as the full 3-sigma potential landing ellipse, present significant challenges.
Mapping or surveying large areas at full resolution becomes massively time consuming, or requires very large teams to do effectively.
One solution is using machine learning systems to classify images by semantic segmentation.
This approach can provide a useful tool to augment the workflow of human geomorphologists, “triaging” extremely large, high resolution datasets to identify concentrations of certain textures or landforms.
I will discuss the process of developing the Oxia planum training dataset, and our assessment of the results.
</p>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
<details>
<summary> <font size="+1"> January 18, 2022 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="image fit captioned">
<h3> Anirudh Koul </br> Pinterest, NASA Frontier Development Lab </h3>
<p> SpaceML Worldview Search: The NoCode Earth & Natural Disaster Dataset Curator from Unlabeled Petabyte Scale Imagery </p>
<p> AI modeling for Earth events at NASA is often limited by the availability of labeled examples.
For example, training classifiers to detect forest fires from satellite imagery requires curating a massive and diverse dataset of example forest fires, a tedious multi-month effort requiring careful review of over 196 million square miles of data per day for 20 years.
While such images might exist in abundance within 40 petabytes of unlabeled satellite data, finding these positive examples to include in a training dataset for a machine learning model is extremely time-consuming and requires researchers to "hunt" for positive examples, like finding a needle in a haystack.
In this presentation, we showcase a no-code open-source tool built by an international team of citizen scientists whose goal is to minimize the amount of human manual image labeling needed to achieve a state-of-the-art classifier.
The pipeline, purpose-built to take advantage of the massive amount of unlabeled images, consists of (1) self-supervision training to convert unlabeled images into meaningful representations, (2) search-by-example to collect a seed set of similar images, (3) human-in-the-loop active learning to iteratively ask for labels on uncertain examples and train on them.
In initial experiments, the system has yielded orders of magnitude reduction in time and cost of data labeling efforts and has shown the potential to multiply the efficiency of the researcher's data curation efforts. </p>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> February 22, 2022 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Doğacan Su Öztürk </br> University of Alaska: Fairbanks </h3>
<p> Towards a predictive high-resolution model of high-latitude ionospheric convection using SuperDARN Radars </p>
<p> Investigations of the meso-scale structures garnered significant interest in recent years due to their significant contribution to the global energy budget of the Ionosphere-Thermosphere system.
Perturbations at such scales are more difficult to predict due to their intricate driver dependencies and the higher spatiotemporal resolutions required to model their behaviour.
In this study, we leverage the seven-year grid data from the Prince George Radar which is operated by the University of Saskatchewan to predict latitudinal and longitudinal ionospheric convection velocities.
This talk will focus on the development of the prediction model and the lessons learnt that will be used towards building a global multi-scale prediction model of ionospheric convection.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Nithin Sivadas </br> NASA Goddard Space Flight Center </h3>
<p> Uncertainty in solar wind propagation leads to biased estimates of the ionospheric response </p>
<p> Solar wind measurements are mainly made at the L1 libration point, 230 Earth radii upstream from the Earth.
To study their impact on the Earth, we need to estimate conditions closer to Earth.
The best estimates of solar wind propagation times from L1 to Earth have an uncertainty of about 10 to 50 minutes.
Also, solar wind features have scale lengths ranging from 30 to 80 RE transverse to the Sun-Earth line.
Shocks and rotational discontinuities propagate through the solar wind, changing its structure as a function of time.
We show that these uncertainties lead to a surprising non-linear bias in the regression slopes of ionospheric response with solar wind drivers.
Data-based modelers need to be aware of how this bias affects the interpretation of their results and the coupling of their model with physics-based models.
Finally, we demonstrate that we can correct this bias using Bayesian techniques and prior knowledge of the uncertainties.
</p>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
<details>
<summary> <font size="+1"> April 12, 2022 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Bharat Kunduri </br> Virginia Tech </h3>
<p> An examination of HF propagation modes using machine learning guided by ray-tracing </p>
<p> The Super Dual Auroral Radar Network (SuperDARN) is a network of HF radars that are typically used for monitoring plasma convection in the Earth's ionosphere.
Due to the complex nature of HF propagation, the radars also observe backscatter from other sources such as ground and meteor trails.
However, it is not straightforward to distinguish between these different modes of backscatter observed due to the similarities in measured parameters.
In this presentation, I will discuss a new machine learning algorithm that is guided by ray-tracing for identifying different backscatter modes in SuperDARN.
The new model can distinguish between meteor scatter, E-/F-region ionospheric, and ground scatter.
The model is validated by comparing predicted elevation angles with those measured at the radar, and the differences between these quantities can further be used to quantify the uncertainties in ionospheric parameters such as NmF2. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Ameya Daigavane </br> Google Research India </h3>
<p> Unsupervised Detection of Magnetic Field Boundary Crossings for Responsive Instrumentation </p>
<p> We investigate unsupervised time-series analysis techniques for real-time identification of magnetic field boundaries in data collected by the Cassini Plasma Spectrometer (CAPS) on the Cassini mission to Saturn.
Previous research has sought to identify these boundaries either in a supervised setting, where a subset of labeled boundary crossing events are available to the algorithms, or using magnetometer data in which these boundaries are significantly clearer.
We find that Saturn bow shock transitions can be reliably detected by the methods we consider, while magnetopause transitions are harder to identify across different years, indicating that generalization of algorithm parameters remains a challenge.
We identify the Plasma Instrument for Magnetic Sounding (PIMS) on the upcoming Europa Clipper Mission as a promising beneficiary of the use of similar processing methods. </p>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> May 24, 2022 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Corey Cochrane </br> NASA Jet Propulsion Laboratory </h3>
<p> Single-Pass Subsurface Ocean Detection at Triton using the Principal Components of Magnetic Induction Measurements </p>
<p> There are many moons in the solar system thought to potentially harbor hidden oceans; Neptune's moon of Triton is a prime example.
Magnetic induction detection is very promising due to the presence of Neptune's strong and highly dynamic magnetic field.
One complication is the confounding presence of an intense ionosphere which can also generate an induced magnetic field that can mask the induction response from the putative ocean.
Additionally, due to the uncertainty of Neptune's rotation rate, the phase of the synodic magnetic wave will be unknown until arrival. In this presentation, we demonstrate an ocean detection methodology based on principal component analysis that is resilient to these factors, in addition to various other noise sources.
We also show how constrained ocean characterization can be achieved through a multi-pass mission concept. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> David Fouhey </br> University of Michigan </h3>
<p> Synthesizing Magnetograms of the Sun's Photosphere with Deep Learning </br> - </p>
<p> The Sun's magnetic field is the source of many high-energy space weather events like solar flares and so it is monitored by a set of instruments, each of which offers trade-offs between cadence, spatial and spectral resolution, and field of view.
For instance, Hinode/SOT-SP is extremely accurate but has a slow cadence and limited field of view, and SDO/HMI is less spectrally capable but produces fast cadence, full-disk data.
Conventionally, data from each instrument is used in isolation to infer the state of the Sun's photospheric magnetic field by inverting a forward model that maps magnetic field conditions to expected polarized light.
I'll show SynthIA, a deep-learning based system that links data from multiple instruments to produce synthetic magnetograms capturing the best aspects of each instrument.
In particular, I'll show that SynthIA can synthesize magnetograms with the cadence and field of view of SDO/HMI as well as the quality resembling those of the spectrally capable Hinode/SOT-SP pipeline. </p>
</div>
</div>
</div>
</div>
</details>
<details>
<summary> <font size="+1"> June 28, 2022 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Don Hood </br> Baylor University </h3>
<p> Automated Boulder Detection on the Martian Surface with MBARS </p>
<p> Boulder-sized clasts are common on the surface of Mars, and many are sufficiently large to be resolved by the High Resolution Imaging Science Experiment (HiRISE) camera aboard the Mars Reconnaissance Orbiter (MRO).
The size, number, and location of boulders on the surface and their spatial distribution can reveal the processes that have operated on the surface, including boulder erosion, burial, impact excavation, and other mechanisms of boulder transport and generation.
However, quantitative analysis of statistically significant boulder populations which could inform these processes entails prohibitively laborious manual segmentation, granulometry and morphometry measurements over large areas.
We have developed an automated tool to locate and measure boulders on the martian surface: the Martian Boulder Automatic Recognition System (MBARS).
The open-source Python-based toolkit autonomously measures boulder diameter and height in HiRISE images enabling rapid and accurate assessments of boulder populations. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Yu Tao </br> University College London </h3>
<p> Super-3D: Subpixel-Scale Topography Retrieval of Mars Using Deep Learning </p>
<p> High-resolution digital terrain models (DTMs) play an important role in studying the formation processes involved in generating a modern-day planetary surface such as Mars.
However, it has been a common understanding that DTMs derived from a particular imaging dataset can only achieve a lower, or at the best, similar effective spatial resolution compared to the input images, due to the various approximations and/or filtering processes introduced by the photogrammetric and/or photoclinometric pipelines.
With recent successes in deep learning techniques, it is now become feasible to improve the effective resolution of an image using super-resolution restoration (SRR) networks, retrieving pixel-scale topography using single-image DTM estimation (SDE) networks, and subsequently, combine the two techniques to produce subpixel-scale topography from only a single-view input image.
In this work, we demonstrate with the <a href = "https://www.mdpi.com/2072-4292/13/9/1777">UCL MARSGAN SRR</a> and the <a href = "https://www.mdpi.com/2072-4292/13/21/4220">MADNet SDE</a> systems to produce single-input-image-based DTMs at <a href = "https://www.mdpi.com/2072-4292/14/2/257">subpixel-scale spatial resolution</a> using the 4 m/pixel ESA Trace Gas Orbiter Colour and Stereo Surface Imaging System (CaSSIS) “PAN” band images and the 25 cm/pixel NASA Mars Reconnaissance Orbiter High Resolution Imaging Science Experiment (HiRISE) “RED” band images.
Co-authors: J-P. Muller (UCL) S. J. Conway (CNRS) </p>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
<!-- January, skipped December. -->
</details>
<details>
<summary> <font size="+1"> <b> SEMINAR SERIES 2020 - 2021 </b> </font>
<p> <a href="https://www.youtube.com/playlist?list=PL4kOs9mVzcpb8vA4gh08wEAPCZayIR_Pq"> Recordings </a> </p>
</summary>
<!-- September -->
<details>
<summary> <font size="+1"> September 22, 2020 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3>Abigail Azari & Caitriona Jackman </br> ML4PSP Organizers</h3>
<p>Introductions to the ML4PSP Series & </br> Integrating ML for Planetary Science In the Next Decade <a href="https://arxiv.org/abs/2007.15129">[paper]</a> </p>
<p>The ML4PSP organizers will discuss the series and provide introductions before summarizing a recent white paper submitted to the NRC Planetary and Astrobiology decadal on integrating machine learning into planetary science. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3>Matthew K. James </br> University of Leicester</h3>
<p>3D modelling of Mercury's magnetosphere </br> using the new MESSENGER FIPS proton moments <a href="https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2019JA027352">[paper]</a></p>
<p>A new MESSENGER FIPS dataset is introduced, where the 𝜅-distribution function is fitted numerically to proton spectra, providing more accurate estimates of density and temperature than previous Maxwellian fits. The quality of the fitted distribution functions are then assessed using modular artificial neural networks in order to remove badly fitted spectra. The new moments are then used to train a deep artificial neural network in order to create a scalable 3D proton model for Mercury's magnetosphere.</p>
</div>
</div>
</div>
</div>
</details>
<!-- October -->
<details>
<summary> <font size="+1"> October 27, 2020 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://www.wkiri.com/"> Kiri Wagstaff </a></br> NASA Jet Propulsion Laboratory / </br> Oregon State University</h3>
<p> Machine Learning for Spacecraft at Europa: Enabling In-Situ Discoveries to Maximize Science Return <a href="https://wkiri.com/research/papers/wagstaff-onboard-europa-19.pdf">[2019 paper]</a>, <a href="https://wkiri.com/research/papers/daigavane-transitions-MiLeTS-20.pdf">[2020 paper]</a></p>
<p>Upcoming missions to remote destinations like Jupiter's moon Europa will operate at extreme distances from the Earth where direct human oversight is impossible. The combination of extreme distance, limited lifetime due to high radiation, and limited data downlink creates an urgent need for reliable autonomous operations. Machine learning can help by analyzing data for features of interest as it is collected. Data with positive detections can be marked for high-priority downlink to Earth for mission planning. For Europa, such features include thermal anomalies, active icy plumes, and unusual surface mineral deposits. This talk describes data analysis and machine learning methods that can operate onboard to increase the rate of exploration and discovery.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://mypages.unh.edu/argallmr/bio"> Matthew Argall </a> </br> University of New Hampshire </br> - </h3>
<p>The MMS SITL Ground Loop: Automating the Burst Selection Process <a href="https://doi.org/10.3389/fspas.2020.00054">[paper]</a>, <a href="https://helioml.org/08/Automated_Detection_Of_Magnetopause_Crossings.html">[book chapter]</a></p>
<p>Global-scale energy flow throughout Earth's magnetosphere is catalyzed by processes that occur at Earth's magnetopause (MP). Magnetic reconnection is one process responsible for solar wind entry into and global convection within the magnetosphere, and the MP location, orientation, and motion have an impact on the dynamics. Statistical studies that focus on these and other MP phenomena and characteristics inherently require MP identification in their event search criteria, a task that can be automated using machine learning. We introduce a Long-Short Term Memory (LSTM) Recurrent Neural Network model into the operational data stream of the Magnetospheric Multiscale (MMS) mission to free up mission operation costs, detect MP crossings, and assist studies of energy transfer into the magnetosphere. </p>
</div>
</div>
</div>
</div>
</details>
<!-- November -->
<details>
<summary> <font size="+1"> November 24, 2020 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Kiley Yeakel </br>The Johns Hopkins University </br> Applied Physics Laboratory</h3>
<p> Machine Learning Algorithms for Automated Detection of Boundary Crossings: A Case Study from Cassini </p>
<p>As increasingly data-intensive sensors are developed for downlink-constrained deep-space missions, scientists face a future in which only a small portion of the science data collected by the spacecraft can be sent back to Earth. There’s a rapidly increasing need to develop “smart” autonomous algorithms capable of rudimentary science analysis on-board the spacecraft so that the downlink bandwidth can be optimized for the most relevant observations. Here, we present one such case study from the Cassini mission where we have utilized machine learning (ML) algorithms to classify whether the spacecraft was in the magnetosphere, magnetosheath or solar wind, utilizing a set of labeled magnetopause and bow shock crossings spanning from 2004 – 2016. We analyze the overall accuracy of various ML algorithms – Recurrent Neural Networks (RNNs) and Gaussian Mixture Models (GMMs) – utilizing combinations of features from the magnetometer (MAG), Charge Mass Spectrometer (CHEMS) and Low-Energy Magnetospheric Measurement System (LEMMS).</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://mariomorvan.github.io/"> Mario Morvan </a></br> University College London </br> - </h3>
<p>Using Deep Learning for Precision Photometry in Exoplanetary Science <a href="https://iopscience.iop.org/article/10.3847/1538-3881/ab6aa7">[2020a paper]</a>, <a href="https://arxiv.org/abs/2011.02030">[2020b preprint]</a></p>
<p>Disentangling the planetary signal from the stellar and instrumental noise is a major data and modelling challenge with inevitable repercussions for transits detection and characterisation. Here we consider approaches to leverage the power of deep learning and help tackling this challenge. After discussing a LSTM-based method to model the noise in Spitzer light curve observations, we present preliminary studies aiming at developing an end-to-end differentiable pipeline combining the flexibility and scalability of neural networks with the precision and domain knowledge borne by physical models.</p>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
<!-- January, skipped December. -->
<details>
<summary> <font size="+1"> January 26, 2021 </font>
</summary>
<p>Special Theme: Model Metrics and Assessment </p>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://clasp.engin.umich.edu/people/michael-liemohn/"> Michael Liemohn </a></br> University of Michigan</h3>
<p>One is Not Enough</br>Thoughts on Choosing Data-Model Comparison Metrics</p>
<p>The magnetospheric physics research community uses a broad array of quantitative data-model comparison methods – metrics – when conducting their research investigations. It is often the case, though, that any particular study will only use one or two metrics. Because metrics are designed to test a specific aspect of the data-model relationship, limiting the comparison to only one or two metrics reduces the physical insights that can be gleaned from the analysis, restricting the possible findings from such studies. Additional physical insights can be obtained when many types of metrics are applied. A few best practices for choosing metrics for space physics studies are presented and discussed.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://www.drsophiemurray.com/"> Sophie Murray </a></br> Dublin Institute of Advanced Studies</h3>
<p>Finding the Right Metric</br>Solar Flare Forecast Evaluation <a href="https://arxiv.org/abs/1907.02905">[paper]</a>, <a href="https://www.swsc-journal.org/articles/swsc/full_html/2020/01/swsc200004/swsc200004.html">[paper]</a>, <a href="https://ccmc.gsfc.nasa.gov/challenges/flare.php">[flare scoreboard]</a></p>
<p>One essential component of operational space weather forecasting is the prediction of solar flares. Early flare forecasting work focused on statistical methods based on historical flaring rates, and more complex machine learning based methods have been implemented in recent years. A multitude of flare forecasting methods are now available for operational use, and proper evaluation of these products is crucially important for model developers, forecasters, end-users, and stakeholders because it facilitates an understanding of the strengths and weaknesses of the forecasting process. This talk will outline current collaborative efforts in solar flare forecasting that are driving international standards based on terrestrial weather forecasting practices, such as defining evaluation metrics, climatological benchmarking, and ensemble requirements.</p>
</div>
</div>
</div>
</div>
</details>
<!-- February -->
<details>
<summary> <font size="+1"> February 23, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://www.dias.ie/2019/07/24/tadhg-garton/"> Tadhg Garton </a></br> University of Southampton /</br> Alan Turing Institute</h3>
<p>Machine Learning identification of signatures in 1D magnetospheric timeseries</p>
<p>The products of magnetic reconnection in Saturn's magnetotail are identified in magnetometer observations primarily through characteristic deviations in the north-south component of the magnetic field. Identification of these features has long been performed by human observers, however with the advent of sophisticated computational methods, it is time to automate our search for these reconnection signatures. Here, we present a fully automated, supervised learning, feed forward neural network model to identify evidence of reconnection in the Kronian magnetosphere with the three magnetic field components observed by the Cassini spacecraft in Kronocentric radial-theta-phi (KRTP) coordinates as input. Furthermore, we present methods to validate results of machine learning algorithms when they are applied to extended datasets that originate in differing background environments than those trained, tested and validated against.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://www.liorruba.com/"> Lior Rubanenko </a></br> Stanford University </br> - </h3>
<p>Automatic detection of barchan dunes on Mars employing an instance segmentation neural network</p>
<p>The surface of Mars is riddled with dunes created by accumulating sand particles that are carried by the wind. When the sand supply is limited and the wind is approximately unidirectional, dunes take the form of crescents termed barchan dunes, whose slip faces are oriented in the dominant wind direction. Consequently, analyzing the morphometrics of barchan dunes can help characterize the winds that form them. Previously, local circulation patterns were derived by analyzing individual images of barchan dunes near the North Pole of Mars <a href="https://doi.org/10.1029/JB084iB14p08167">[1]</a>. However, repeating this analysis on a global scale remains a challenge, as manually mapping dunes is largely impractical and traditional computer vision algorithms are largely ineffective at identifying the outlines of dunes from images. Here we employ Mask R-CNN <a href="https://arxiv.org/abs/1703.06870">[2]</a>, an instance segmentation convolutional neural network, to map dunes across the surface of Mars. Training on ~1000 images, our model achieves a mean average detection precision (mAP) of 80%, for IoU = 0.5. In the talk, I will describe the Mask R-CNN neural network and its vast space of hyperparameters, and how those can be employed for object detection and analysis by incorporating traditional computer vision techniques.</p>
</div>
</div>
</div>
</div>
</details>
<!-- March -->
<details>
<summary> <font size="+1"> March 23, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://www.mn.uio.no/fysikk/english/people/aca/lbnc/"> Lasse Clausen </a></br> University of Oslo </h3>
<p>Auroral images: Automatic classification and geomagnetic predictions <a href="https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018JA025274">[paper]</a></p>
<p>We use a pre-trained deep neural network to automatically extract features from auroral images. Using a manually labelled training dataset, we are then able to automatically classify images into one of the following classes: "clear", "cloudy", "moon", "arc", "diffuse", and "discrete". As a next step, we show some initial results from our attempts to use the extracted features to predict magnetometer observations.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://hannah-rae.github.io/"> Hannah Kerner </a></br> University of Maryland </h3>
<p>Novelty-Guided Target Selection for Mars Rovers</p>
<p>Mars rover operations currently consists of pre-scripted commands determined by the rover science and engineering operations teams on a day-to-day (or sol-to-sol) basis. Automated instrument targeting systems, which determine which surface features to target based on automated rather than pre-scripted decision-making, could help increase the science return from current and future exploration missions. To enable automatic follow-up observation of novel targets—i.e., targets that differ substantially from those observed previously in the mission—we propose to use novelty detection algorithms for ranking candidate targets detected in rover images. In this talk, I will present our proposed onboard novelty detection framework and illustrate its utility using diverse scenarios of novel geology found in Mars rover images.</p>
</div>
</div>
</div>
</div>
</details>
<!-- April -->
<details>
<summary> <font size="+1"> April 27, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="https://dept.atmos.ucla.edu/jbortnik"> Jacob Bortnik </a></br> University of California, Los Angeles </h3>
<p> Machine learning reconstruction of the inner magnetosphere </p>
<p> While the volumes of space physics data continue to rise exponentially, our analysis techniques have not kept pace with this rapid growth, and often do not exploit the full potential of the data. In this talk, we discuss how machine learning might provide the solution to this problem, and in particular we will show how a (sparse) time-series of point observations of some quantity can be converted into a 3-dimensional time-varying model of that quantity with the use of neural networks. As an example, we show a three-dimensional dynamic electron density (DEN3D) model in the inner magnetosphere, that can provide full coverage of the inner magnetosphere and in fact is sufficiently accurate that it points the way to new physical discoveries. The talk will be concluded with a few emerging ideas of how machine learning can be applied in the physical sciences. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Téo Bloch </br> University of Reading </h3>
<p> Deep-Ensemble Modelling of Electron Flux at the Radiation Belt’s Outer Boundary With Bayesian Neural Networks <a href="https://sites.google.com/view/rbflux-bnn">[website]</a></p>
<p> As space-based infrastructure becomes more ubiquitous, modelling the radiation belts is increasingly important. Most radiation belt models require an accurate outer boundary condition, as this helps to drives the simulation. Our work aims to characterise the flux-energy distribution at the outer boundary location using an ensemble of Bayesian neural networks. Each model in the ensemble predicts 11 values of flux, and the associated variance, for each set of inputs. The model performs well, predicting fluxes within a factor of 2.5 for the lower energies and within a factor of 4 for the higher energies (with a correlation between 0.5-0.8). </p>
</div>
</div>
</div>
</div>
</details>
<!-- May -->
<details>
<summary> <font size="+1"> May 25, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3><a href="http://ryanmcgranaghan.com/"> Ryan McGranaghan </a></br> Atmospheric and Space Technology Research Associates (ASTRA) </h3>
<p> Advancing space physics research through machine learning and information representation: A powerful and demonstrable use case through solar wind-magnetosphere-ionosphere coupling </p>
<p>The connection between the Sun and the Earth is a complex one, involving interactions and variabilities across a dizzying spectrum of scales and systems.
The result is a relationship between us and our star that is observable only through a fleet of instruments, methods, and technologies yet creates weather in the near-Earth space environment colloquially known as space weather.
Space weather is the impact of solar energy on society and is a powerful use case for demonstrating the ability of data science to converge disciplines.
A key to understanding it is the way that regions of space between the Sun and the Earth’s surface are connected, particularly via particles transferred from the solar wind to the magnetosphere to the upper atmosphere—a problem that remains one of the great challenges in space physics.
We will first present a new ML model that better captures the dynamics of the particle precipitation from a large volume of data.
We will then share a new framework to evaluate and understand these models.
We will generalize this progress as suggestive of trends that reverberate across all scientific disciplines and that even tie science to engineering, art, and design.
We will raise generative questions about how we currently do AI/ML research in space physics, about the role of information representation in progressing space physics discovery (e.g., knowledge networks), and to provide insight to and spark discussion for this cross-disciplinarity community around the concepts of convergence and antidisciplinary.</p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Giacomo Nodjoumi </br> Jacobs University </br> Bremen </h3>
<p> Detecting Cave Entrance Candidates on Mars using Deep Learning Computer Vision </p>
<p>In the framework of geological exploration of terrestrial planets, sinkhole-like landforms (pit craters, pit chains and skylights), as a potential direct access to the subsurface, are one of the most promising environments to focus our research.
Detecting, mapping, and describing those types of landforms is a challenging process since a set of tedious tasks must be conducted manually by researchers, usually on a small set of available data.
These tasks vary from data collection (in which areas with high probability of occurrences are selected and downloaded) to manual analysis that requires viewing the images in detail, mapping all occurrences with GIS, and extracting morphometric parameters.
For the Moon and Mars, databases of cave candidates exist (see MGC^3 Cushing, 2012) but all of these databases are focused on small regions, rather than at planetary scale.
Thus there are possible missing correlations between the presence of these landforms and the area of detection that can be related to past and present processes.
To achieve data analyses at planetary scale, machine learning and deep learning algorithms are extremely valuables techniques, capable of automatically analysing large datasets.
The main problem is that often it is necessary to develop specific tools and pipelines for this task.
Two jupyter notebooks have been developed around FOSS Object Detection packages and used with a small dataset of 130 high resolution images acquired by the HiRISE camera from the Mars Reconnaissance Orbiter.
The aim of these is to create an user-friendly environment for training and evaluating an object detection model; not only for cave candidates but also for other types of landforms.
These results have been compared to available databases with preliminary promising results.</p>
</div>
</div>
</div>
</div>
</details>
<!-- June -->
<details>
<summary> <font size="+1"> June 29, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> Victor Pinto </br> University of New Hampshire </h3>
<p> Reproducibility in Space Sciences </br> Do we really need to publish our codes? </p>
<p> Reproducibility is without a doubt one of the fundamental pillars of science, and just as science has evolved through time, so has the concept of what is necessary to ensure results are reproducible. Historically, the publication of source code used for analysis, modeling, and even for figures has never been a requirement in Space Sciences and it has rarely been encouraged as a necessary practice. A possible explanation may be the assumption that reproducibility is “baked in” the physics of the systems or in the data. However, as Machine Learning becomes an integral part of Space Sciences research, the requirement that the codes developed for research are publicly available, or at least can be obtained on demand grows in popularity. Today, many publishers are encouraging code sharing, or even straight up requiring it now or in the near future. This talk will stray away from the traditional discussions on techniques and applications of machine learning for space and planetary sciences and focus on the topics of whether code sharing is the best practice to ensure reproducibility, with the idea of starting a conversation on whether we should push for or against code sharing as a community, and how should we prepare for it if it becomes the norm moving forward. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> Stéphane Aicardi </br> Observatoire de Paris </h3>
<p> Deep Learning on Jovian Decametric Emissions </br> - </p>
<p> Jupiter's decametric radio emissions have long been a valuable tool for understanding the magnetic properties of Jupiter and its interactions with its major satellites. Decametric emissions are variable on many timescales, but high-resolution observations produce huge data that can't be stored for long. Using convolutional neural networks, we try to detect, classify and locate Jovian emissions in the observations by the Nançay decameter array. This will lead to an automated method to select the high resolution data to be archived. See predictions from this network <a href="https://voparis-minerva-jupiter.obspm.fr/">online</a>. </p>
</div>
</div>
</div>
</div>
</details>
<!-- July -->
<details>
<summary> <font size="+1"> July 27, 2021 </font>
</summary>
<div class="container">
<div class="row 300%">
<div class="6u 12u$(xsmall)">
<div class="image fit captioned">
<h3> John Biersteker </br> Massachusetts Institute of Technology </br> - </h3>
<p> Probing Europa's interior with Bayesian inference and magnetic induction </p>
<p> Exploring subsurface oceans on icy moons is a key goal in the search for habitable environments.
Because these moons are embedded in the time-varying magnetosphere of their host planet, their saltwater oceans generate an induced magnetic field detectable by spacecraft.
Such measurements provide a tool to detect and characterize these ocean worlds.
As part of the design of the upcoming Europa Clipper mission, we have developed a technique for ocean characterization using spacecraft magnetometry and Bayesian inference.
This approach allows for the recovery of the ocean parameters with robust uncertainties and enables incorporating multiple datasets into a self-consistent view of Europa's interior.
I will discuss the application of this technique to the upcoming Europa Clipper mission, archival data from Galileo, and possible future missions to the ice giants. </p>
</div>
</div>
<div class="6u$ 12u$(xsmall)">
<div class="image fit captioned">
<h3> <a href="https://cires.colorado.edu/cires-members-council-researcher/dr-hazel-m-bain">Hazel Bain</a> </br> CIRES University of Colorado Boulder </br> NOAA Space Weather Prediction Center </h3>
<p> How Well Can We Forecast Solar Radiation Storms? </p>
<p> Solar energetic particles (SEPs) are a driver of space weather, the effects of which can impact high-frequency communications systems, satellite systems and pose a radiation hazard for astronauts, as well flight crew and passengers on polar flight routes.
The National Oceanic and Atmospheric Administration's Space Weather Prediction Center (NOAA/SWPC) issues space weather forecasts and products for energetic protons at Earth.
I will discuss a recent verification study to assess our ability to forecast these storms.
It is hoped that this study will serve as a benchmark for the development and validation of physics-based and machine learning SEP models. Co-authors: R. Steenburgh (NOAA SWPC), T. Onsager (NOAA SWPC), E. M. Stitely (Millersville University) </p>
</div>
</div>
</div>
</div>
</details>
<!-- <font size="+1"> - </font> -->
</details>
</section>
<!-- Footer -->
<footer id="footer">
<div class="container">
<ul class="icons">
<li><a href="https://www.youtube.com/channel/UCyS8UU1a6K0NM_lSN7ykAnA/playlists" class="icon fa-youtube"></a></li>
<!--<li><a href="#" class="icon fa-twitter"></a></li>-->
<!--<li><a href="#" class="icon fa-instagram"></a></li>-->
</ul>
<ul class="copyright">
<li>© ML4PSP Organizers </li>
<!--<li>Design: <a href="https://templated.co/spatial"> TEMPLATED </a></li> -->
<li>Design: TEMPLATED </li>
<!--<li>Images: <a href="http://unsplash.com">Unsplash</a></li>-->
</ul>
</div>
</footer>
<!-- Scripts -->
<script src="assets/js/jquery.min.js"></script>
<script src="assets/js/skel.min.js"></script>
<script src="assets/js/util.js"></script>
<script src="assets/js/main.js"></script>
</body>
</html>