Using NeRFs and 3DGS for 4D representation generation and reconstruction, enhancing dynamic scene understanding.

Mingrui Zhao, Sauradip Nag, Kai Wang et al.

VFM-VAE leverages frozen VFM as encoder with novel decoder, achieving gFID 1.62 in 640 epochs, 10× faster than prior methods.

Tianci Bi, Xiaoyi Zhang, Yan Lu et al.

Glyph renders long texts into images processed by VLMs, achieving 3-4× token compression with performance comparable to Qwen3-8B.

Jiale Cheng, Yusen Liu, Xinyu Zhang et al.

Proposes a context-aware pseudo-label scoring framework for zero-shot video summarization, achieving 57.58 F1 on SumMe.

Yuanli Wu, Long Zhang, Yue Du et al.

VISTA employs multi-agent self-improvement, achieving up to 60% win rate in video quality enhancement.

Do Xuan Long, Xingchen Wan, Hootan Nakhost et al.

XModBench evaluates Gemini 2.5 Pro's cross-modal consistency, revealing less than 60% accuracy in spatial and temporal reasoning.

Xingrui Wang, Jiang Liu, Chao Huang et al.

Proposes NEO, a native vision-language model trained on 390M image-text pairs, achieving pixel-word alignment and surpassing modular models in various benchmarks.

Haiwen Diao, Mingxuan Li, Silei Wu et al.

WithAnyone model achieves controllable and ID-consistent image generation using contrastive loss and the MultiID-2M dataset.

Hengyuan Xu, Wei Cheng, Peng Xing et al.

Proposes Generative Universal Verifier, trained on ViVerBench, improving visual verification by 8.3 points with OmniVerifier-7B and TTS strategies.

Xinchen Zhang, Xiaoying Zhang, Youbin Wu et al.

FusionNet and Transformer-based models achieved top PSNR of 26.35 in NTIRE 2025 low-light enhancement, demonstrating multi-model fusion effectiveness.

Xiaoning Liu, Zongwei Wu, Florin-Alexandru Vasluianu et al.

Mask-GRPO introduces reinforcement learning into masked generative models, significantly improving text-to-image generation with a 0.73 score on GenEval and FID of 8.32, surpassing SOTA.

Yifu Luo, Xinhao Hu, Keyu Fan et al.

DepthVLA integrates a pretrained depth module into a mixture-of-transformers framework, significantly improving spatial reasoning and manipulation success rates (e.g., 78.5% in real-world tasks).

Tianyuan Yuan, Yicheng Liu, Chenhao Lu et al.

EgoSocial leverages multimodal cues to improve proactive intervention detection in OLMMs, boosting timing accuracy by 45.6% on Phi-4.

Xijun Wang, Tanay Sharma, Achin Kulshrestha et al.

DriveVLA-W0 employs world modeling to predict future images, providing dense self-supervision that significantly enhances autonomous driving models' scalability, outperforming baselines on large datasets.

Yingyan Li, Shuyao Shang, Weisong Liu et al.

AnyUp is a universal feature upsampling method that generalizes to any feature type at inference, outperforming state-of-the-art.

Thomas Wimmer, Prune Truong, Marie-Julie Rakotosaona et al.

DrivingScene employs a static-to-dynamic two-stage training with residual scene flow, achieving real-time high-fidelity 4D scene reconstruction from two images.

Qirui Hou, Wenzhang Sun, Chang Zeng et al.

Proposes a novel ML framework for detecting corner cases in autonomous driving, enhancing detection accuracy.

Sebastian Schmidt, Julius Körner, Stephan Günnemann

Proposes a domain-conditioned, image-free instance generation method that boosts ILR performance by 15% on average across seven benchmarks.

Yankun Wu, Zakaria Laskar, Giorgos Kordopatis-Zilos et al.

FlexTraj introduces a point-based trajectory control framework enabling multi-granularity, alignment-agnostic image-to-video synthesis with improved efficiency.

Zhiyuan Zhang, Can Wang, Dongdong Chen et al.

Proposed a guided diffusion model with Transformer for hyperspectral data augmentation, boosting forest classification accuracy by 4.8%.

Mattia Ferrari, Lorenzo Bruzzone