Keras Applications provides a collection of pre‑trained deep‑learning models for image classification, feature extraction, and transfer learning. Below is a curated list of 20 alternative libraries, model zoos, and platforms that offer comparable or complementary pre‑trained vision models for image‑recognition tasks.
Get targeted exposure with custom position pinning and highlighted placement.
Repository of reusable, pre‑trained TensorFlow models (including image classification, object detection, and segmentation) that can be fine‑tuned or used as feature extractors.
Official hub for PyTorch offering a wide range of pre‑trained vision models (ResNet, EfficientNet, YOLO, etc.) with simple one‑line loading.
Part of the TorchVision library, it provides state‑of‑the‑art image classification, detection, and segmentation models ready for PyTorch workflows.
A collection of pre‑trained models in the Open Neural Network Exchange (ONNX) format, compatible with many frameworks and optimized for inference.
High‑level API built on PyTorch that includes ready‑to‑use image classification and segmentation models with easy transfer‑learning utilities.
Apache MXNet’s computer‑vision toolkit offering a rich set of pre‑trained models for classification, detection, and segmentation.
Supports loading pre‑trained models from Caffe, TensorFlow, ONNX, and Torch, enabling fast inference directly within OpenCV.
Legacy collection of pre‑trained Caffe models for image classification and detection, still useful for lightweight deployments.
Offers a variety of pre‑trained vision models (ResNet, MobileNet, SSD, etc.) optimized for MXNet’s hybrid execution.
Baidu’s deep‑learning platform with a curated set of pre‑trained image models for classification, detection, and segmentation.
Growing hub of vision‑specific transformer models (ViT, DeiT, Swin, CLIP) with ready‑to‑use pipelines for image classification and retrieval.
Enterprise‑grade repository of GPU‑optimized pre‑trained models (including EfficientNet, ResNet, YOLO, and segmentation nets) for inference on NVIDIA hardware.
Marketplace of reusable ML assets, including TensorFlow and PyTorch vision models, with one‑click deployment to Vertex AI.
Microsoft’s collection of pre‑trained models and notebooks, featuring vision models that integrate seamlessly with Azure Machine Learning.
Curated set of pre‑trained vision models (e.g., ResNet, EfficientNet, YOLO) that can be deployed directly on SageMaker with minimal code.
Intel’s repository of optimized pre‑trained models for edge inference, covering classification, detection, and segmentation with OpenVINO runtime.
Platform for sharing and discovering pre‑trained object detection and classification models, with automatic dataset preprocessing pipelines.
Open‑source server for deploying deep learning models (Caffe, TensorFlow, PyTorch) with REST/GRPC APIs, includes several pre‑trained vision models.
Catalog of pre‑trained models (including image classification, object detection, and style transfer) packaged as Docker containers for easy deployment.
Differentiable computer vision library for PyTorch that also provides a set of pre‑trained models and utilities for image processing pipelines.