Research map / Vision-language and multimodal / Human and embodied modeling
DualToken: Towards Unifying Visual Understanding and Generation with Dual Visual Vocabularies
Wei Song, Yuran Wang, Zijia Song, Yadong Li, Zenan Zhou, Long Chen, Xu Jhua, Jiaqi Wang, Kaicheng Yu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| train encoder for reconstruction only instead of text-aligned (SigLIP) pretraining (ablation) | LLaVA-1.5 vision encoder · MMBench | accuracy: 61.8 -> 36.2 | Hurt |
| apply recon and semantic losses on the same deep layer (single vocabulary) instead of hierarchical (ablation) | DualToken tokenizer · ImageNet-1K | rFID / zero-shot: rFID 0.25 -> 3.86; zero-shot 82.0 -> 72.3 | Hurt |
| apply reconstruction loss to shallow layer 6 and semantic loss to deep layer 26 (ablation) | DualToken tokenizer · ImageNet-1K | rFID: 3.86 -> 0.25 | Improved |
| add semantic loss to a reconstruction-only shallow tokenizer (ablation) | DualToken tokenizer · ImageNet-1K | rFID: 0.29 -> 0.25 | No clear effect |
| feed semantic plus pixel tokens instead of semantic tokens only (ablation) | DualToken in LLaVA-1.5 · ten understanding benchmarks | average score: 52.2 -> 53.9 | Improved |
| replace VILA-U tokenizer with DualToken in LLaVA-1.5 | LLaVA-1.5 · ten understanding benchmarks | average score: 48.1 -> 53.9 | Improved |
| generate semantic tokens alongside pixel tokens instead of pixel tokens only (ablation) | DualToken-3B · GenAI-Bench advanced | overall VQAScore: 0.55 -> 0.68 | Improved |
| use one unified tokenizer instead of separate MoVQGAN + SigLIP dual encoders (ablation) | DualToken-3B · MJHQ-30K | FID: 17.55 -> 7.88 | Improved |
| use DualToken 3B instead of VILA-U 3B tokenizer | unified AR model · MJHQ-30K | FID: 15.12 -> 7.88 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- UniTok — UniTok: a Unified Tokenizer for Visual Generation and Understanding (NeurIPS 2025) · better on 2 of 2 shared results
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
- Vision as a Dialect: Unifying Visual Understanding and Generation via Text-Aligned Representations (NeurIPS 2025 · similarity 0.91)
- TokenFlow: Unified Image Tokenizer for Multimodal Understanding and Generation (CVPR 2025 · similarity 0.91)
- UniFlow: A Unified Pixel Flow Tokenizer for Visual Understanding and Generation (ICLR 2026 · similarity 0.91)
- UniTok: a Unified Tokenizer for Visual Generation and Understanding (NeurIPS 2025 · similarity 0.90)
- Rethinking Visual Token Reduction in LVLMs Under Cross-Modal Misalignment (AAAI 2026 · similarity 0.90)
- G$^2$TR: Generation-Guided Visual Token Reduction for Separate-Encoder Unified Multimodal Models (NeurIPS 2026 · similarity 0.90)
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.
