Scaling Image Tokenizers with Grouped Spherical Quantization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Jiangtao, Qin, Zhen, Zhang, Yifan, Hu, Vincent Tao, Ommer, Björn, Briq, Rania, Kesselheim, Stefan
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917856378093568
author Wang, Jiangtao
Qin, Zhen
Zhang, Yifan
Hu, Vincent Tao
Ommer, Björn
Briq, Rania
Kesselheim, Stefan
author_facet Wang, Jiangtao
Qin, Zhen
Zhang, Yifan
Hu, Vincent Tao
Ommer, Björn
Briq, Rania
Kesselheim, Stefan
contents Vision tokenizers have gained a lot of attraction due to their scalability and compactness; previous works depend on old-school GAN-based hyperparameters, biased comparisons, and a lack of comprehensive analysis of the scaling behaviours. To tackle those issues, we introduce Grouped Spherical Quantization (GSQ), featuring spherical codebook initialization and lookup regularization to constrain codebook latent to a spherical surface. Our empirical analysis of image tokenizer training strategies demonstrates that GSQ-GAN achieves superior reconstruction quality over state-of-the-art methods with fewer training iterations, providing a solid foundation for scaling studies. Building on this, we systematically examine the scaling behaviours of GSQ, specifically in latent dimensionality, codebook size, and compression ratios, and their impact on model performance. Our findings reveal distinct behaviours at high and low spatial compression levels, underscoring challenges in representing high-dimensional latent spaces. We show that GSQ can restructure high-dimensional latent into compact, low-dimensional spaces, thus enabling efficient scaling with improved quality. As a result, GSQ-GAN achieves a 16x down-sampling with a reconstruction FID (rFID) of 0.50.
format Preprint
id arxiv_https___arxiv_org_abs_2412_02632
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Scaling Image Tokenizers with Grouped Spherical Quantization
Wang, Jiangtao
Qin, Zhen
Zhang, Yifan
Hu, Vincent Tao
Ommer, Björn
Briq, Rania
Kesselheim, Stefan
Computer Vision and Pattern Recognition
Artificial Intelligence
Vision tokenizers have gained a lot of attraction due to their scalability and compactness; previous works depend on old-school GAN-based hyperparameters, biased comparisons, and a lack of comprehensive analysis of the scaling behaviours. To tackle those issues, we introduce Grouped Spherical Quantization (GSQ), featuring spherical codebook initialization and lookup regularization to constrain codebook latent to a spherical surface. Our empirical analysis of image tokenizer training strategies demonstrates that GSQ-GAN achieves superior reconstruction quality over state-of-the-art methods with fewer training iterations, providing a solid foundation for scaling studies. Building on this, we systematically examine the scaling behaviours of GSQ, specifically in latent dimensionality, codebook size, and compression ratios, and their impact on model performance. Our findings reveal distinct behaviours at high and low spatial compression levels, underscoring challenges in representing high-dimensional latent spaces. We show that GSQ can restructure high-dimensional latent into compact, low-dimensional spaces, thus enabling efficient scaling with improved quality. As a result, GSQ-GAN achieves a 16x down-sampling with a reconstruction FID (rFID) of 0.50.
title Scaling Image Tokenizers with Grouped Spherical Quantization
topic Computer Vision and Pattern Recognition
Artificial Intelligence
url https://arxiv.org/abs/2412.02632