AIQViT: Architecture-Informed Post-Training Quantization for Vision Transformers

TL;DR

AIQViT introduces a novel post-training quantization method for vision transformers that reduces information loss and improves performance by using architecture-informed compensation and dynamic focusing quantization.

Contribution

The paper proposes AIQViT, a new PTQ approach that incorporates low-rank compensation and dynamic quantization for better ViT performance.

Findings

Outperforms state-of-the-art PTQ methods on five vision tasks.

Effectively reduces performance degradation in low-bit quantization.

Enhances quantization of post-Softmax activations with dynamic focus.

Abstract

Post-training quantization (PTQ) has emerged as a promising solution for reducing the storage and computational cost of vision transformers (ViTs). Recent advances primarily target at crafting quantizers to deal with peculiar activations characterized by ViTs. However, most existing methods underestimate the information loss incurred by weight quantization, resulting in significant performance deterioration, particularly in low-bit cases. Furthermore, a common practice in quantizing post-Softmax activations of ViTs is to employ logarithmic transformations, which unfortunately prioritize less informative values around zero. This approach introduces additional redundancies, ultimately leading to suboptimal quantization efficacy. To handle these, this paper proposes an innovative PTQ method tailored for ViTs, termed AIQViT (Architecture-Informed Post-training Quantization for ViTs). First, we design an architecture-informed low rank compensation mechanism, wherein learnable low-rank weights are introduced to compensate for the degradation caused by weight quantization. Second, we design a dynamic focusing quantizer to accommodate the unbalanced distribution of post-Softmax activations, which dynamically selects the most valuable interval for higher quantization resolution. Extensive experiments on five vision tasks, including image classification, object detection, instance segmentation, point cloud classification, and point cloud part segmentation, demonstrate the superiority of AIQViT over state-of-the-art PTQ methods.