AI Paper Atlas

Research map / Video / Video-language understanding

SweetTok: Semantic-Aware Spatial-Temporal Tokenizer for Compact Video Discretization

Zhentao Tan, Ben Xue, Jian Jia, Junhao Wang, Wencai Ye, Shaoyun Shi, Mingjie Sun, Wenjin Wu, Quan Chen, Peng Jiang

ICCV 2025 · Video › Video-language understanding · approach: Long video understanding · 7 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
decouple spatial and temporal queries (DQAE) instead of vanilla query-based compression (ablation)SweetTok · UCF-101 reconstructionrFVD: 35.15 -> 20.46Improved
use decoupled queries instead of naive 1D linear-interpolation downsampling to 1280 tokens (ablation)SweetTok · UCF-101 reconstructionrFVD: 227.65 -> 20.46Improved
add vanilla language codebook (ablation)SweetTok · UCF-101 reconstructionrFVD: 29.45 -> 24.80Improved
make the language codebook motion-enhanced (verbs/adverbs for temporal queries) (ablation)SweetTok · UCF-101 reconstructionrFVD: 24.80 -> 20.46Improved
replace CLIP embeddings with larger Qwen embeddings in the MLC (ablation)SweetTok · UCF-101 reconstructionrFVD: 20.46 -> 20.12No clear effect
remove token compression (5,120 tokens instead of 1,280)SweetTok · UCF-101 reconstructionrFVD: 10.74 on UCF-101 and 7.51 on K-600Improved
scale the AR generator from 650M to 1.9B parametersSweetTok · UCF-101 generationgFVD: 84 -> 65Improved
use SweetTok tokens instead of OmniTok tokens for generationVideoGPT AR generator · UCF-101 generationgFVD: 191 -> 84Improved
use SweetTok tokens instead of LARP tokens for generationAR generator · UCF-101 generationgFVD: 99 -> 84Improved
fine-tune SweetTok spatial branch on ImageNet with 1,024 tokensSweetTok · ImageNet reconstructionrFID: 0.59 -> 0.37Improved

Later papers that compared with this method

1 later accepted paper reports SweetTok as a baseline in their results tables; 0 of them report a better number on most of the results they share with it.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy