Wavelet-Based Image Tokenizer for Vision Transformers

• URL: http://arxiv.org/abs/2405.18616v1
• Date: Tue, 28 May 2024 21:45:46 GMT
• Title: Wavelet-Based Image Tokenizer for Vision Transformers
• Authors: Zhenhai Zhu, Radu Soricut,
• Abstract summary: Non-overlapping patch-wise convolution is the default image tokenizer for vision Transformer (ViT) models. We propose a new image tokenizer based on wavelet transformation. We show that ViT models with the new tokenizer achieve both higher training throughput and better top-1 precision for the ImageNet validation set.
• Score: 14.346214923088855
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Non-overlapping patch-wise convolution is the default image tokenizer for all state-of-the-art vision Transformer (ViT) models. Even though many ViT variants have been proposed to improve its efficiency and accuracy, little research on improving the image tokenizer itself has been reported in the literature. In this paper, we propose a new image tokenizer based on wavelet transformation. We show that ViT models with the new tokenizer achieve both higher training throughput and better top-1 precision for the ImageNet validation set. We present a theoretical analysis on why the proposed tokenizer improves the training throughput without any change to ViT model architecture. Our analysis suggests that the new tokenizer can effectively handle high-resolution images and is naturally resistant to adversarial attack. Furthermore, the proposed image tokenizer offers a fresh perspective on important new research directions for ViT-based model design, such as image tokens on a non-uniform grid for image understanding.

Related papers

• Joint Post-Training Quantization of Vision Transformers with Learned Prompt-Guided Data Generation [5.75627633588113]
  We present a framework for joint quantization of Vision Transformers trained on ImageNet.<n>We achieve state-of-the-art W4A4 and W3A3 accuracies on ImageNet.<n>We also introduce a data-free calibration strategy that synthesizes diverse, label-free samples.
  arXiv  Detail & Related papers  (2026-02-21T15:02:21Z)
• ViT-FIQA: Assessing Face Image Quality using Vision Transformers [18.848638277203616]
  Face Image Quality Assessment (FIQA) aims to predict the utility of a face image for face recognition (FR) systems.<n>ViT-FIQA is a novel approach that extends standard ViT backbones, originally optimized for FR, through a learnable quality token.<n>Experiments on challenging benchmarks and several FR models, including both CNN- and ViT-based architectures, demonstrate that ViT-FIQA consistently achieves top-tier performance.
  arXiv  Detail & Related papers  (2025-08-19T15:50:07Z)
• Towards Generalizable AI-Generated Image Detection via Image-Adaptive Prompt Learning [33.269644831847636]
  Image-Adaptive Prompt Learning (IAPL) is a novel paradigm that adjusts the prompts according to each input image, rather than fixing them after training.<n>IAPL achieves state-of-the-art performance, with mean accuracies of 95.61% and 96.7% on the widely used UniversalFakeDetect and GenImage datasets.
  arXiv  Detail & Related papers  (2025-08-03T05:41:24Z)
• Vision Foundation Models as Effective Visual Tokenizers for Autoregressive Image Generation [52.261584726401686]
  We present a novel direction to build an image tokenizer directly on top of a frozen vision foundation model.<n>Based on these designs, our proposed image tokenizer, VFMTok, achieves substantial improvements in image reconstruction and generation quality.
  arXiv  Detail & Related papers  (2025-07-11T09:32:45Z)
• End-to-End Vision Tokenizer Tuning [73.3065542220568]
  The vision tokenizer optimized for low-level reconstruction is to downstream tasks requiring varied representations and semantics.<n>The loss of the vision tokenization can be the representation bottleneck for target tasks.<n>We propose ETT, an end-to-end vision tokenizer tuning approach that enables joint optimization between vision tokenization and target autoregressive tasks.
  arXiv  Detail & Related papers  (2025-05-15T17:59:39Z)
• Make A Long Image Short: Adaptive Token Length for Vision Transformers [5.723085628967456]
  We propose an innovative approach to accelerate the ViT model by shortening long images. Specifically, we introduce a method for adaptively assigning token length for each image at test time to accelerate inference speed.
  arXiv  Detail & Related papers  (2023-07-05T08:10:17Z)
• Visual Prompt Tuning for Generative Transfer Learning [26.895321693202284]
  We present a recipe for learning vision transformers by generative knowledge transfer. We base our framework on state-of-the-art generative vision transformers that represent an image as a sequence of visual tokens to the autoregressive or non-autoregressive transformers. To adapt to a new domain, we employ prompt tuning, which prepends learnable tokens called prompt to the image token sequence.
  arXiv  Detail & Related papers  (2022-10-03T14:56:05Z)
• Image and Model Transformation with Secret Key for Vision Transformer [16.055655429920993]
  We show for the first time that models trained with plain images can be directly transformed to models trained with encrypted images. The performance of the transformed models is the same as models trained with plain images when using test images encrypted with the key.
  arXiv  Detail & Related papers  (2022-07-12T08:02:47Z)
• Modeling Image Composition for Complex Scene Generation [77.10533862854706]
  We present a method that achieves state-of-the-art results on layout-to-image generation tasks. After compressing RGB images into patch tokens, we propose the Transformer with Focal Attention (TwFA) for exploring dependencies of object-to-object, object-to-patch and patch-to-patch.
  arXiv  Detail & Related papers  (2022-06-02T08:34:25Z)
• MAT: Mask-Aware Transformer for Large Hole Image Inpainting [79.67039090195527]
  We present a novel model for large hole inpainting, which unifies the merits of transformers and convolutions. Experiments demonstrate the state-of-the-art performance of the new model on multiple benchmark datasets.
  arXiv  Detail & Related papers  (2022-03-29T06:36:17Z)
• Towards a Unified Foundation Model: Jointly Pre-Training Transformers on Unpaired Images and Text [93.11954811297652]
  We design a unified transformer consisting of modality-specific tokenizers, a shared transformer encoder, and task-specific output heads. We employ the separately-trained BERT and ViT models as teachers and apply knowledge distillation to provide additional, accurate supervision signals. Experiments show that the resultant unified foundation transformer works surprisingly well on both the vision-only and text-only tasks.
  arXiv  Detail & Related papers  (2021-12-14T00:20:55Z)
• Vector-quantized Image Modeling with Improved VQGAN [93.8443646643864]
  We propose a Vector-quantized Image Modeling approach that involves pretraining a Transformer to predict image tokens autoregressively. We first propose multiple improvements over vanilla VQGAN from architecture to codebook learning, yielding better efficiency and reconstruction fidelity. When trained on ImageNet at 256x256 resolution, we achieve Inception Score (IS) of 175.1 and Frechet Inception Distance (FID) of 4.17, a dramatic improvement over the vanilla VQGAN.
  arXiv  Detail & Related papers  (2021-10-09T18:36:00Z)
• Visual Saliency Transformer [127.33678448761599]
  We develop a novel unified model based on a pure transformer, Visual Saliency Transformer (VST), for both RGB and RGB-D salient object detection (SOD) It takes image patches as inputs and leverages the transformer to propagate global contexts among image patches. Experimental results show that our model outperforms existing state-of-the-art results on both RGB and RGB-D SOD benchmark datasets.
  arXiv  Detail & Related papers  (2021-04-25T08:24:06Z)
• So-ViT: Mind Visual Tokens for Vision Transformer [27.243241133304785]
  We propose a new classification paradigm, where the second-order, cross-covariance pooling of visual tokens is combined with class token for final classification. We develop a light-weight, hierarchical module based on off-the-shelf convolutions for visual token embedding. The results show our models, when trained from scratch, outperform the competing ViT variants, while being on par with or better than state-of-the-art CNN models.
  arXiv  Detail & Related papers  (2021-04-22T09:05:09Z)
• Tokens-to-Token ViT: Training Vision Transformers from Scratch on ImageNet [128.96032932640364]
  We propose a new Tokens-To-Token Vision Transformers (T2T-ViT) to solve vision tasks. T2T-ViT reduces the parameter counts and MACs of vanilla ViT by 200%, while achieving more than 2.5% improvement when trained from scratch on ImageNet. For example, T2T-ViT with ResNet50 comparable size can achieve 80.7% top-1 accuracy on ImageNet.
  arXiv  Detail & Related papers  (2021-01-28T13:25:28Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.