AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

SPAR: Semantic-Pixel Self-Alignment and Adaptive Routing for Unified Multimodal Models

Hongxiang Li, Hongxu chen, chenyang zhu, Xiaoshuang Huang, Jiayin Cai, Xiaolong Jiang, Yao Hu, Long Chen

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Instruction-tuned multimodal assistants · 1 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
add pixel stream to the tokenizer (ablation)SPAR · ImageNet 50k reconstructionPSNR: 24.62 -> 26.65Improved
add pixel stream to the tokenizer (ablation)SPAR · ImageNet 50k reconstructionSSIM: 0.788 -> 0.856Improved
add pixel stream to the tokenizer (ablation)SPAR · MMBenchaccuracy: 72.6 -> 73.0No clear effect
add semantic stream to the tokenizer (ablation)SPAR · MMBenchaccuracy: 18.4 -> 73.0Improved
add semantic stream to the tokenizer (ablation)SPAR · MME-Pscore: 709 -> 1500Improved
add semantic stream to the tokenizer (ablation)SPAR · ImageNet 50k reconstructionPSNR: 25.28 -> 26.65Improved
unfreeze encoder with dual-stream training instead of freezing (ablation)SPAR · ImageNet 50k reconstructionrFID: 6.14 -> 0.27Improved
unfreeze encoder with dual-stream training instead of freezing (ablation)SPAR · MMBenchaccuracy: 72.6 -> 73.0Improved
add Dynamic Token Routing instead of last-layer extraction (ablation)SPAR · GenEvaloverall: 0.86 -> 0.89Improved
add Dynamic Token Routing instead of last-layer extraction (ablation)SPAR · ImgEditoverall: 3.73 -> 3.85Improved
add self-alignment loss (ablation)SPAR · GenEvaloverall: 0.86 -> 0.89Improved
add self-alignment loss (ablation)SPAR · WISEoverall: 0.53 -> 0.57Improved

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.

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