# DeepCABAC: A Universal Compression Algorithm for Deep Neural Networks

arXiv: https://arxiv.org/abs/1907.11900 · HTML: https://mlsanity.com/paper/1907.11900/

## Key facts

- Cited by 26 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 27 Jul 2019.

## Built on by (most cited first, 26 of 26)

- [An Information-Theoretic Justification for Model Pruning](https://mlsanity.com/paper/2102.08329/) (2021)
- [ECQx: Explainability-Driven Quantization for Low-Bit and Sparse DNNs](https://mlsanity.com/paper/2109.04236/) (2021)
- [End-to-End Learnable Multi-Scale Feature Compression for VCM](https://mlsanity.com/paper/2306.16670/) (2023)
- [Prune Your Model Before Distill It](https://mlsanity.com/paper/2109.14960/) (2021)
- [Neural Network Compression for Noisy Storage Devices](https://mlsanity.com/paper/2102.07725/) (2021)
- [A Privacy Preserving System for Movie Recommendations Using Federated Learning](https://mlsanity.com/paper/2303.04689/) (2023)
- [Point Cloud Compression with Implicit Neural Representations: A Unified Framework](https://mlsanity.com/paper/2405.11493/) (2024)
- [Communication-Efficient Federated Learning via Predictive Coding](https://mlsanity.com/paper/2108.00918/) (2021)
- [Low Rank Optimization for Efficient Deep Learning: Making A Balance between Compact Architecture and Fast Training](https://mlsanity.com/paper/2303.13635/) (2023)
- [Can We Find Strong Lottery Tickets in Generative Models?](https://mlsanity.com/paper/2212.08311/) (2022)
- [Lossy and Lossless (L2) Post-training Model Size Compression](https://mlsanity.com/paper/2308.04269/) (2023)
- [Approximate Bisimulation Relations for Neural Networks and Application to Assured Neural Network Compression](https://mlsanity.com/paper/2202.01214/) (2022)
- [Binarized Neural Networks Converge Toward Algorithmic Simplicity: Empirical Support for the Learning-as-Compression Hypothesis](https://mlsanity.com/paper/2505.20646/) (2025)
- [Reward-Based 1-bit Compressed Federated Distillation on Blockchain](https://mlsanity.com/paper/2106.14265/) (2021)
- [Safety Verification of Neural Network Control Systems Using Guaranteed Neural Network Model Reduction](https://mlsanity.com/paper/2301.07531/) (2023)
- [Semi-Supervised Learning for Image Classification using Compact Networks in the BioMedical Context](https://mlsanity.com/paper/2205.09678/) (2022)
- [Implicit Neural Compression of Point Clouds](https://mlsanity.com/paper/2412.10433/) (2024)
- [AnyPcc: Compressing Any Point Cloud with a Single Universal Model](https://mlsanity.com/paper/2510.20331/) (2025)
- [Compressing AI Traffic: Standardized Neural Network Coding of Visual-Token Representations in Split Vision-Language Inference](https://mlsanity.com/paper/2609.01200/) (2026)
- [AirNet: Neural Network Transmission over the Air](https://mlsanity.com/paper/2105.11166/) (2021)
- [Frequency-Aware Re-Parameterization for Over-Fitting Based Image Compression](https://mlsanity.com/paper/2310.08068/) (2023)
- [Compression Repair for Feedforward Neural Networks Based on Model Equivalence Evaluation](https://mlsanity.com/paper/2402.11737/) (2024)
- [Evaluation of strategies for efficient rate-distortion NeRF streaming](https://mlsanity.com/paper/2410.19459/) (2024)
- [Reducing Storage of Pretrained Neural Networks by Rate-Constrained Quantization and Entropy Coding](https://mlsanity.com/paper/2505.18758/) (2025)
- [P2U: Progressive Precision Update For Efficient Model Distribution](https://mlsanity.com/paper/2506.22871/) (2025)
- [D-Quant: Driftable Entropy Coding for KV Cache Quantization](https://mlsanity.com/paper/2609.19880/) (2026)

## Abstract (arXiv)

The field of video compression has developed some of the most sophisticated and efficient compression algorithms known in the literature, enabling very high compressibility for little loss of information. Whilst some of these techniques are domain specific, many of their underlying principles are universal in that they can be adapted and applied for compressing different types of data. In this work we present DeepCABAC, a compression algorithm for deep neural networks that is based on one of the state-of-the-art video coding techniques. Concretely, it applies a Context-based Adaptive Binary Arithmetic Coder (CABAC) to the network's parameters, which was originally designed for the H.264/AVC video coding standard and became the state-of-the-art for lossless compression. Moreover, DeepCABAC employs a novel quantization scheme that minimizes the rate-distortion function while simultaneously taking the impact of quantization onto the accuracy of the network into account. Experimental results show that DeepCABAC consistently attains higher compression rates than previously proposed coding techniques for neural network compression. For instance, it is able to compress the VGG16 ImageNet model by x63.6 with no loss of accuracy, thus being able to represent the entire network with merely 8.7MB. The source code for encoding and decoding can be found at https://github.com/fraunhoferhhi/DeepCABAC.