CVLGJul 12, 2022

Synergistic Self-supervised and Quantization Learning

arXiv:2207.05432v116 citationsh-index: 21Has Code
Originality Highly original
AI Analysis

This addresses the deployment challenge of self-supervised models in resource-constrained settings, offering a novel method to enhance quantization robustness without extra storage.

The paper tackles the problem of severe accuracy drops in self-supervised learning models when quantized for resource-constrained applications, proposing SSQL to pretrain quantization-friendly models that improve accuracy at lower bit-widths and boost full precision performance in most cases.

With the success of self-supervised learning (SSL), it has become a mainstream paradigm to fine-tune from self-supervised pretrained models to boost the performance on downstream tasks. However, we find that current SSL models suffer severe accuracy drops when performing low-bit quantization, prohibiting their deployment in resource-constrained applications. In this paper, we propose a method called synergistic self-supervised and quantization learning (SSQL) to pretrain quantization-friendly self-supervised models facilitating downstream deployment. SSQL contrasts the features of the quantized and full precision models in a self-supervised fashion, where the bit-width for the quantized model is randomly selected in each step. SSQL not only significantly improves the accuracy when quantized to lower bit-widths, but also boosts the accuracy of full precision models in most cases. By only training once, SSQL can then benefit various downstream tasks at different bit-widths simultaneously. Moreover, the bit-width flexibility is achieved without additional storage overhead, requiring only one copy of weights during training and inference. We theoretically analyze the optimization process of SSQL, and conduct exhaustive experiments on various benchmarks to further demonstrate the effectiveness of our method. Our code is available at https://github.com/megvii-research/SSQL-ECCV2022.

Code Implementations1 repo
Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes