Next Forcing introduces multi-chunk prediction to accelerate training and improve accuracy in high-frame-rate video generation, achieving 94.1% success on RoboTwin.

Gangwei Xu, Qihang Zhang, Jiaming Zhou et al.

AMNet introduces modality-agnostic inference for low-light video enhancement, maintaining high performance even with missing auxiliary modalities, outperforming state-of-the-art methods.

Hangfeng Liang, Yutao Hu, Yanhan Hu et al.

Introduces Data2Story, a multi-agent framework transforming data into verifiable multimodal stories with evidence traceability and interactive content.

Kevin Qinghong Lin, Batu EI, Yuhong Shi et al.

MOFA-VTON employs diffusion models with dual-region masks and cross-attention-based layout adjustment, enabling user-controlled, fine-grained virtual try-on with diverse styles.

Xiaoyu Han, Chenyang Wang, Jing Wang et al.

POTATR is a lightweight 29M-parameter image-to-graph model that significantly improves page-level table extraction accuracy and efficiency.

Brandon Smock, Libin Liang, Max Sokolov et al.

TEVI leverages sparse autoencoders with text conditioning to refine image embeddings, significantly improving vision-language alignment and retrieval accuracy.

Sweta Mahajan, Sukrut Rao, Jiahao Xie et al.

This paper introduces a unified framework based on watching, remembering, and reasoning, significantly advancing long video understanding with multimodal LLMs.

Jiahao Meng, Yue Tan, Qi Xu et al.

PAR3D introduces part-aware 3D multimodal large language models, significantly enhancing fine-grained scene understanding via the ScenePart dataset.

Shaohui Dai, Yansong Qu, You Shen et al.

Proposes Complexity-Balanced Diffusion Splitting (CBS), using Dirichlet energy and trajectory acceleration to estimate local complexity, improving synthesis quality by ~35%.

Noam Issachar, Dani Lischinski, Raanan Fattal

Proposes Astra framework combining RL-trained VLM policy with Bagel-based world simulator for imagination-driven spatial reasoning, improving MMSI-Bench accuracy from 45.1% to 49.5%.

Chenming Zhu, Jingli Lin, Yilin Long et al.

Proposes MedReCo, an entity-aware vision-language framework with over 690,000 images for clinical case retrieval and change description.

Tengfei Zhang, Ziheng Zhao, Lisong Dai et al.

HomeWorld introduces a hierarchical, multimodal framework trained on 300K real floorplans, using LLMs and diffusion models to generate controllable, diverse, and realistic whole-home scenes.

Wenbo Li, Xiaoliang Ju, Zipeng Qin et al.

Proposes a knowledge refinement framework using automatic rule mining and ASP-based abductive reasoning, improving scene graph generation with +4-8% F1@50 across benchmarks.

Maëlic Neau, Salim Baloch, Jakob Suchan et al.

GMBFormer integrates NDVI-guided global memory with Transformer for urban green-space extraction, achieving a mean IoU of 89.25%.

Hao Lei, Xi Cheng, Chenlu Shu et al.

Proposes PhaseLock, a training-free framework that extracts motion priors from 2-step inference, improving physical consistency by 6.2 points on average.

Woojung Han, Seil Kang, Youngjun Jun et al.

Proposes SEIG, a staged framework leveraging pretrained vision-language models (VLMs) to reconstruct editable 3D scenes from a single image, achieving high fidelity in geometry, materials, and lighting.

Guangzhao He, Rundong Luo, Wei-Chiu Ma et al.

AdaCodec employs predictive visual coding, transmitting full reference frames only when prediction is costly, reducing visual tokens by 84.7% and boosting long-video understanding efficiency.

Haowen Hou, Zhen Huang, Zheming Liang et al.

This paper introduces VLM as a teacher for video reasoning via test-time online optimization, achieving a 16.7-point performance boost, surpassing traditional methods.

Junhao Cheng, Liang Hou, Tianxiong Zhong et al.

Lumos-Nexus employs a two-stage training and UPFB to bridge frequencies, boosting video fidelity and reasoning-driven generation.

Jiazheng Xing, Hangjie Yuan, Lingling Cai et al.

Introduces SOCO benchmark with over 1 million keypoint pairs across 100 categories, revealing that vision foundation models encode strong semantics but poorly transfer correspondences across categories.

Olaf Dünkel, Basavaraj Sunagad, Haoran Wang et al.