BranchVision-KerasModels 🚀\n\n\n\n\n\n## 📌 Table of Contents\n- 📖 Introduction\n- ⚡ Installation\n- 🛠️ Usage\n- 📑 Models\n- 📜 License\n- 🌟 Credits\n\n## 📖 Introduction\nBranchVision-KerasModels is a refined library of vision models with pretrained weights, optimized for object detection, segmentation, vision-language modeling (VLMs), and more with backbone flexibility, such as in1k, in21k, and others.\n\n## ⚡ Installation\nInstall via PyPI:\nshell\npip install -U branchvision\n\n\nInstall from source:\nshell\npip install -U git+https://github.com/branchonly/BranchVision-KerasModels\n\n\n## 🛠️ Usage\n### 🔎 Listing Available Models\npython\nimport branchvision as kvmm\nprint(kvmm.list_models())\n\n\n### 🔎 Listing Specific Model Variant\npython\nimport kvmm\nprint(kvmm.list_models("swin"))\n\n\n### ⚙️ Using Layers\npython\nimport kvmm\nlayer = kvmm.layers.StochasticDepth(drop_path_rate=0.1)\noutput = layer(input_tensor, training=True)\n\n\n### 🏗️ Backbone Usage\npython\nimport kvmm\nmodel = kvmm.models.vit.ViTTiny16()\n\n\n### Example Inference\npython\nfrom keras.applications.imagenet_utils import decode_predictions\nimport kvmm\nmodel = kvmm.models.swin.SwinTinyP4W7(input_shape=[224, 224, 3])\nimage = Image.open("bird.png").resize((224, 224))\npreds = model.predict(ops.expand_dims(ops.convert_to_tensor(image), axis=0))\nprint("Predicted:", decode_predictions(preds, top=3)[0])\n\n\n## 📑 Models\nExplore backbones, segmentation, and VLM implementations with associated research papers and weight sources.\n\n### 🔖 License\nLicensed under Apache 2.0.\n\n## 🌟 Credits\nAcknowledgements to Keras, Transformers, and timm communities for their contributions to ML ecosystems.\n\n## Citing\nbash\n@misc{branchvision2025models,\n author = {BranchVision Team},\n title = {BranchVision-KerasModels},\n year = {2025},\n publisher = {GitHub},\n journal = {GitHub repository},\n howpublished = {\url{https://github.com/branchonly/BranchVision-KerasModels}}\n}\n