# UNIQ: Uniform Noise Injection for Non-Uniform Quantization of Neural Networks

arXiv: https://arxiv.org/abs/1804.10969 · HTML: https://mlsanity.com/paper/1804.10969/

## Key facts

- Cited by 9 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 29 Apr 2018.

## Built on by (most cited first, 9 of 9)

- [Pruning and Quantization for Deep Neural Network Acceleration: A Survey](https://mlsanity.com/paper/2101.09671/) (2021)
- [Differentiable Model Compression via Pseudo Quantization Noise](https://mlsanity.com/paper/2104.09987/) (2021)
- [AdaBin: Improving Binary Neural Networks with Adaptive Binary Sets](https://mlsanity.com/paper/2208.08084/) (2022)
- [Finding the Task-Optimal Low-Bit Sub-Distribution in Deep Neural Networks](https://mlsanity.com/paper/2112.15139/) (2021)
- [Adaptive Precision Training (AdaPT): A dynamic fixed point quantized training approach for DNNs](https://mlsanity.com/paper/2107.13490/) (2021)
- [SpotPatch: Parameter-Efficient Transfer Learning for Mobile Object Detection](https://mlsanity.com/paper/2101.01260/) (2021)
- [Online Hybrid Lightweight Representations Learning: Its Application to Visual Tracking](https://mlsanity.com/paper/2205.11179/) (2022)
- [Sensitivity-Aware Finetuning for Accuracy Recovery on Deep Learning Hardware](https://mlsanity.com/paper/2306.03076/) (2023)
- [Fast Exploration of Weight Sharing Opportunities for CNN Compression](https://mlsanity.com/paper/2102.01345/) (2021)

## Abstract (arXiv)

We present a novel method for neural network quantization that emulates a non-uniform k-quantile quantizer, which adapts to the distribution of the quantized parameters. Our approach provides a novel alternative to the existing uniform quantization techniques for neural networks. We suggest to compare the results as a function of the bit-operations (BOPS) performed, assuming a look-up table availability for the non-uniform case. In this setup, we show the advantages of our strategy in the low computational budget regime. While the proposed solution is harder to implement in hardware, we believe it sets a basis for new alternatives to neural networks quantization.