# fchollet/deep-learning-models

**Attribution required: if you use, quote, or summarise this content, you must credit and link back to [awesome-repositories.com](https://awesome-repositories.com/repository/fchollet-deep-learning-models).**

7,349 stars · 2,436 forks · Python · MIT

## Links

- GitHub: https://github.com/fchollet/deep-learning-models
- awesome-repositories: https://awesome-repositories.com/repository/fchollet-deep-learning-models.md

## Description

This project is a collection of deep learning tools for image classification and audio tagging, providing a repository of pre-trained model weights and architectures. It serves as a Keras model zoo that enables the immediate use of established neural networks for inference and transfer learning.

The library includes a music tagging framework that classifies audio recordings using convolutional recurrent neural networks and mel-spectrograms. For visual data, it provides implementations of architectures such as ResNet, VGG, and Xception, alongside a repository of weights trained on large datasets like ImageNet.

The project covers a broad range of capabilities including computer vision and audio analysis. It supports the generation of visual feature maps through layer-based feature extraction and provides workflows for adapting pre-existing networks to new datasets.

## Tags

### Artificial Intelligence & ML

- [Image Classification](https://awesome-repositories.com/f/artificial-intelligence-ml/image-classification.md) — Implements a comprehensive set of convolutional neural networks for categorizing visual data. ([source](https://github.com/fchollet/deep-learning-models/search))
- [Keras Model Implementations](https://awesome-repositories.com/f/artificial-intelligence-ml/neural-network-implementations/keras-model-implementations.md) — Provides a collection of neural network architectures implemented specifically using the Keras API.
- [Mel-Spectrogram Processing](https://awesome-repositories.com/f/artificial-intelligence-ml/audio-processing/mel-spectrogram-processing.md) — Transforms raw audio waveforms into mel-spectrograms before passing them into convolutional neural networks.
- [Image Classification Models](https://awesome-repositories.com/f/artificial-intelligence-ml/computer-vision-systems/computer-vision/image-classification-models.md) — Utilizes pre-trained Keras models to identify objects in images without requiring training from scratch. ([source](https://github.com/fchollet/deep-learning-models/blob/master/README.md))
- [Deep Learning Architectures](https://awesome-repositories.com/f/artificial-intelligence-ml/deep-learning-architectures.md) — Implements deep learning architectures like ResNet and VGG for complex visual recognition.
- [Pre-trained Model Checkpoints](https://awesome-repositories.com/f/artificial-intelligence-ml/large-scale-model-training/vision-transformer-pre-training/pre-trained-model-checkpoints.md) — Includes a library of weights trained on ImageNet to accelerate the development of custom vision models.
- [Hybrid Convolutional Recurrent Networks](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/frameworks/model-construction/neural-network-layers/convolution-layers/hybrid-convolutional-recurrent-networks.md) — Implements hybrid convolutional recurrent networks to process temporal audio data for music tagging.
- [Pre-trained Model Zoos](https://awesome-repositories.com/f/artificial-intelligence-ml/machine-learning/infrastructure/model-training-and-tuning/fine-tuning-and-customization/model-fine-tuning/pre-trained-model-zoos.md) — Serves as a model zoo providing pre-trained architectures and weights for immediate inference.
- [Music Audio Tagging](https://awesome-repositories.com/f/artificial-intelligence-ml/music-audio-tagging.md) — Categorizes audio recordings by processing mel-spectrograms through convolutional recurrent neural networks. ([source](https://github.com/fchollet/deep-learning-models/blob/master/music_tagger_crnn.py))
- [Audio Tagging Frameworks](https://awesome-repositories.com/f/artificial-intelligence-ml/music-genre-classifiers/audio-tagging-frameworks.md) — Provides a specialized framework for classifying audio recordings using convolutional recurrent neural networks and mel-spectrograms.
- [Transfer Learning](https://awesome-repositories.com/f/artificial-intelligence-ml/transfer-learning.md) — Provides workflows for adapting pre-existing networks to new datasets using ImageNet weights.
- [Convolutional Neural Network Architectures](https://awesome-repositories.com/f/artificial-intelligence-ml/convolutional-neural-network-architectures.md) — Implements the VGG19 architecture for image recognition using pre-trained or random weights. ([source](https://github.com/fchollet/deep-learning-models/blob/master/vgg19.py))
- [ResNet Variants](https://awesome-repositories.com/f/artificial-intelligence-ml/convolutional-neural-network-architectures/resnet-variants.md) — Implements various ResNet architectures for efficient image recognition using residual blocks. ([source](https://github.com/fchollet/deep-learning-models/blob/master/resnet50.py))
- [VGG Implementations](https://awesome-repositories.com/f/artificial-intelligence-ml/convolutional-neural-network-architectures/vgg-implementations.md) — Provides implementations of VGG convolutional neural networks with configurable classification layers. ([source](https://github.com/fchollet/deep-learning-models/blob/master/vgg16.py))
- [Depthwise Separable Convolutions](https://awesome-repositories.com/f/artificial-intelligence-ml/convolutional-neural-networks/pointwise-convolutions/depthwise-separable-convolutions.md) — Utilizes depthwise separable convolutions within its architectures to reduce trainable parameters.
- [Visual Feature Extractors](https://awesome-repositories.com/f/artificial-intelligence-ml/deep-learning-architectures/visual-feature-extractors.md) — Enables the generation of visual feature maps by removing final classification layers from deep networks.
- [Feature Extraction](https://awesome-repositories.com/f/artificial-intelligence-ml/feature-extraction.md) — Extracts visual feature maps by removing the final classification layers of deep networks.
- [Pre-trained Weight Loading](https://awesome-repositories.com/f/artificial-intelligence-ml/large-scale-model-training/vision-transformer-pre-training/pre-trained-model-checkpoints/pre-trained-weight-loading.md) — Implements mechanisms to download and apply serialized weights to models for immediate inference. ([source](https://github.com/fchollet/deep-learning-models#readme))
- [Residual Networks](https://awesome-repositories.com/f/artificial-intelligence-ml/residual-networks.md) — Implements residual networks that use skip connections to prevent gradient degradation in deep models.
- [Pretrained Weight Initializers](https://awesome-repositories.com/f/artificial-intelligence-ml/weight-initialization/pretrained-weight-initializers.md) — Allows initializing models with pre-trained ImageNet weights to improve convergence during transfer learning. ([source](https://github.com/fchollet/deep-learning-models/blob/master/vgg16.py))

### Graphics & Multimedia

- [Audio & Music](https://awesome-repositories.com/f/graphics-multimedia/audio-music.md) — Assigns descriptive tags to music recordings using deep learning architectures. ([source](https://github.com/fchollet/deep-learning-models#readme))
- [Image Feature Extraction](https://awesome-repositories.com/f/graphics-multimedia/image-feature-extraction.md) — Generates high-level visual feature maps by applying pooling to intermediate neural representations. ([source](https://github.com/fchollet/deep-learning-models/blob/master/README.md))

### Part of an Awesome List

- [Image Classification Architectures](https://awesome-repositories.com/f/awesome-lists/ai/image-classification-architectures.md) — Provides various deep learning model architectures designed for image recognition tasks. ([source](https://github.com/fchollet/deep-learning-models#readme))
- [Xception Implementations](https://awesome-repositories.com/f/awesome-lists/ai/vision-architectures/xception-implementations.md) — Provides the Xception architecture for categorizing images into classes. ([source](https://github.com/fchollet/deep-learning-models/blob/master/xception.py))

### DevOps & Infrastructure

- [Neural Layer Extraction](https://awesome-repositories.com/f/devops-infrastructure/oci-layer-extraction/neural-layer-extraction.md) — Allows the isolation of specific internal layers to obtain latent tensor representations for analysis.