TCFormer: Visual Recognition via Token Clustering Transformer
This work addresses a bottleneck in vision transformers for computer vision tasks, offering an incremental improvement by dynamically clustering tokens based on semantics.
The paper tackles the problem of fixed token distribution in vision transformers, which disregards semantic meaning and leads to sub-optimal performance, by proposing TCFormer that generates dynamic vision tokens based on semantic clustering, resulting in improved effectiveness across applications like image classification, human pose estimation, semantic segmentation, and object detection.
Transformers are widely used in computer vision areas and have achieved remarkable success. Most state-of-the-art approaches split images into regular grids and represent each grid region with a vision token. However, fixed token distribution disregards the semantic meaning of different image regions, resulting in sub-optimal performance. To address this issue, we propose the Token Clustering Transformer (TCFormer), which generates dynamic vision tokens based on semantic meaning. Our dynamic tokens possess two crucial characteristics: (1) Representing image regions with similar semantic meanings using the same vision token, even if those regions are not adjacent, and (2) concentrating on regions with valuable details and represent them using fine tokens. Through extensive experimentation across various applications, including image classification, human pose estimation, semantic segmentation, and object detection, we demonstrate the effectiveness of our TCFormer. The code and models for this work are available at https://github.com/zengwang430521/TCFormer.