awesome-repositories.comCategoriesBlog
MCP
awesome-repositories.com

Discover the best open-source repositories with AI-powered search.

ExploreCurated searchesOpen-source alternativesSelf-hosted softwareBlogSitemap
ProjectMCP serverAboutHow we rankPress
LegalPrivacyTerms
© 2026 Bringes Technology SRL·VAT RO45896025·hello@awesome-repositories.com
·
liuzhuang13 avatar

liuzhuang13/DenseNet

0
View on GitHub↗

DenseNet

DenseNet is a computer vision model and convolutional neural network implementation designed for image recognition and classification tasks. It utilizes a densely connected network architecture where each layer is connected to every other layer to improve feature propagation.

The implementation reduces the number of parameters while maintaining accuracy through a dense-connectivity pattern and layer-aggregation concatenation. It supports model construction using both standard and bottleneck-compressed architectures, with configurable network depth and growth rates to balance inference time and memory usage.

The project covers convolutional neural network optimization and deep learning model training for visual tasks. It includes GPU memory optimization techniques, such as shared gradient inputs, to reduce the memory footprint during training.

AI search

Explore more awesome repositories

Describe what you need in plain English — the AI ranks thousands of curated open-source projects by relevance.

Start searching with AI

Features

  • Dense Neural Networks - Implements a densely connected network architecture where every layer connects to all subsequent layers for optimal feature propagation.
  • Computer Vision Models - Provides a deep learning architecture tailored for image recognition and classification tasks.
  • Computer Vision Research - Provides an implementation of DenseNet for evaluating performance on standard image recognition benchmarks.
  • Convolutional Neural Network Architectures - Implements a deep convolutional architecture specifically designed to improve feature reuse for image classification.
  • Growth Rate Controls - Controls the number of feature maps added by each layer to manage the model size and memory usage.
  • CNN Architectures - Provides a complete convolutional neural network implementation based on the DenseNet architecture.
  • Feature Map Concatenations - Combines feature maps from all preceding layers using concatenation instead of summation.
  • CNN Hyperparameter Optimization - Allows tuning of network depth and growth rates to optimize memory usage and inference speed.
  • Model Construction - Enables the construction of densely connected convolutional layers using both standard and bottleneck-compressed designs.
  • Bottleneck Layers - Implements bottleneck layers that use 1x1 convolutions to compress feature maps and reduce computational overhead.
  • Model Width Configurations - Deno DenseNet allows tuning the depth and growth rate of the network to balance inference time, memory usage, and accuracy.
  • PyTorch Implementations - Implements the DenseNet model using the PyTorch framework for image recognition.
  • Training Memory Optimizers - Employs shared gradient inputs to reduce the GPU memory footprint during the backpropagation process.
  • Deep Neural Network Training Optimization - Optimizes the memory footprint during the training of densely connected layers on GPUs.
  • Computer Vision - Densely connected convolutional networks.
  • Image Classification Architectures - Original implementation of densely connected convolutional networks.
4,862 stars·1,067 forks·Lua·BSD-3-Clause·20 views

Star history

Star history chart for liuzhuang13/densenetStar history chart for liuzhuang13/densenet

Frequently asked questions

What does liuzhuang13/densenet do?

DenseNet is a computer vision model and convolutional neural network implementation designed for image recognition and classification tasks. It utilizes a densely connected network architecture where each layer is connected to every other layer to improve feature propagation.

What are the main features of liuzhuang13/densenet?

The main features of liuzhuang13/densenet are: Dense Neural Networks, Computer Vision Models, Computer Vision Research, Convolutional Neural Network Architectures, Growth Rate Controls, CNN Architectures, Feature Map Concatenations, CNN Hyperparameter Optimization.

What are some open-source alternatives to liuzhuang13/densenet?

Open-source alternatives to liuzhuang13/densenet include: afshinea/stanford-cs-230-deep-learning — This repository collects illustrated single-page cheat sheets that compress the core topics of Stanford's CS 230 deep… leoxiaobin/deep-high-resolution-net.pytorch — This project is a PyTorch implementation of a research architecture designed for high-resolution representation… lukemelas/efficientnet-pytorch — This is a PyTorch implementation of EfficientNet convolutional neural networks. It serves as a computer vision model… kaiminghe/deep-residual-networks — This project provides a deep residual network framework and pre-trained PyTorch models designed for high-accuracy… pytorch/vision — This project is a comprehensive computer vision library for the PyTorch ecosystem, providing a standardized collection… cadene/pretrained-models.pytorch — This project is a pretrained model library for PyTorch, providing a collection of convolutional neural network…

Open-source alternatives to DenseNet

Similar open-source projects, ranked by how many features they share with DenseNet.
  • afshinea/stanford-cs-230-deep-learningafshinea avatar

    afshinea/stanford-cs-230-deep-learning

    7,028View on GitHub↗

    This repository collects illustrated single-page cheat sheets that compress the core topics of Stanford's CS 230 deep learning course into visual reference summaries. The collection covers convolutional neural networks, recurrent neural networks, and practical training techniques, pairing schematic diagrams with mathematical notation to bridge intuition and formal understanding. The cheat sheets are organized by subject area and link related concepts across topics, such as connecting vanishing gradients to LSTM gates, to reinforce the full deep learning workflow. Practical training advice on

    cheatsheetconvolutional-neural-networksdata-science
    View on GitHub↗7,028
  • leoxiaobin/deep-high-resolution-net.pytorchleoxiaobin avatar

    leoxiaobin/deep-high-resolution-net.pytorch

    4,479View on GitHub↗

    This project is a PyTorch implementation of a research architecture designed for high-resolution representation learning. It serves as a computer vision framework focused on precise keypoint detection, human pose estimation, and semantic image segmentation. The implementation provides specialized tools for identifying anatomical landmarks on the human body and predicting facial keypoint coordinates to analyze orientation and alignment. It utilizes a system of multi-resolution parallel streams and repeated multi-scale fusion to maintain high-resolution representations throughout the network.

    Cuda
    View on GitHub↗4,479
  • kaiminghe/deep-residual-networksKaimingHe avatar

    KaimingHe/deep-residual-networks

    6,738View on GitHub↗

    This project provides a deep residual network framework and pre-trained PyTorch models designed for high-accuracy image recognition. It implements a neural network architecture that utilizes skip connections to enable the training of very deep models without gradient degradation. The system is designed for computer vision tasks, including image classification, object detection, and visual data segmentation. It includes weights trained on ImageNet to support transfer learning and the fine-tuning of models on custom image datasets. The architectural design focuses on residual learning blocks,

    View on GitHub↗6,738
  • lukemelas/efficientnet-pytorchlukemelas avatar

    lukemelas/EfficientNet-PyTorch

    8,223View on GitHub↗

    This is a PyTorch implementation of EfficientNet convolutional neural networks. It serves as a computer vision model library providing architectures for image classification and high-level feature extraction, including pre-trained weights for immediate image categorization. The library supports transfer learning by allowing the modification of model architectures and output layers to accommodate a custom number of classes for new datasets. It also includes a model exporter to convert trained PyTorch weights into the ONNX format for production inference. The system covers broader computer vis

    Python
    View on GitHub↗8,223
See all 30 alternatives to DenseNet→