Proposes EBiEOT framework combining semi-supervised data via inverse entropic OT for conditional distribution maximization.

Mikhail Persiianov, Arip Asadulaev, Nikita Andreev et al.

Self-eXplainable AI with attention mechanisms enhances transparency in medical image analysis, covering 200 papers.

Junlin Hou, Sicen Liu, Yequan Bie et al.

Introduces Girsanov-based convergence bounds for discrete diffusion models in CTMC framework; KL divergence nearly linear in dimension d.

Zikun Zhang, Zixiang Chen, Quanquan Gu

Gradient-free cutting plane method for deep ReLU networks, providing convergence guarantees and improving sample efficiency in active learning.

Erica Zhang, Fangzhao Zhang, Mert Pilanci

Generates symbolic music using LLM-enhanced MetaScore dataset, providing text and tag controls.

Weihan Xu, Julian McAuley, Taylor Berg-Kirkpatrick et al.

ExACT combines R-MCTS and Exploratory Learning, achieving 6%-30% improvement on VisualWebArena.

Xiao Yu, Baolin Peng, Vineeth Vajipey et al.

FeelAnyForce uses a multi-head Transformer with RGB and depth images to estimate 3D contact forces, achieving 4% MAE on unseen objects.

Amir-Hossein Shahidzadeh, Gabriele Caddeo, Koushik Alapati et al.

LS-HAR leverages language-guided skeleton and visual features with Salient Fusion, achieving 89.5% accuracy on NTU-RGB+D XSub dataset.

Mohammad Mahdavian, Mohammad Loni, Ted Samuelsson et al.

Introduces matrix stochastic heat equation (MSHE), derives explicit invariant measure in 1D, and demonstrates classical integrability in weak-noise regime via inverse scattering.

Alexandre Krajenbrink, Pierre Le Doussal

Leopard model excels in multi-image text tasks, surpassing Llama-3.2 using 1.2M open-source data.

Mengzhao Jia, Wenhao Yu, Kaixin Ma et al.

Proposed training-free Probabilistic Speculative Jacobi Decoding (SJD) accelerates autoregressive text-to-image generation by ~2×, maintaining quality and diversity.

Yao Teng, Han Shi, Xian Liu et al.

CTDN method for text-supervised egocentric semantic segmentation, significantly improving accuracy.

Zhaofeng Shi, Heqian Qiu, Lanxiao Wang et al.

Proposed VideoCLIP-XL with TPCM, achieving significant improvements in long description video understanding, trained on over 2 million video-long description pairs.

Jiapeng Wang, Chengyu Wang, Kunzhe Huang et al.

ManiSkill3 is a GPU-parallelized robotics simulator supporting 12 task domains, achieving 30,000+ FPS with 2-3x less memory, enabling fast visual RL training.

Stone Tao, Fanbo Xiang, Arth Shukla et al.

COGIS employs Gaussian Process Implicit Surfaces to online model obstacles, enhancing robot manipulation in partial observability with constraint-based dataset refinement.

Abhinav Kumar, Peter Mitrano, Dmitry Berenson

DeSTA2 uses automatic speech-text pair generation, avoiding instruction tuning, to enhance speech understanding.

Ke-Han Lu, Zhehuai Chen, Szu-Wei Fu et al.

EndoDepth benchmark evaluates robustness of monocular endoscopic depth prediction using mDERS and SCARED-C dataset, revealing strengths and weaknesses of SOTA models.

Ivan Reyes-Amezcua, Ricardo Espinosa, Christian Daul et al.

UniSumEval evaluates 9 LLMs on fine-grained, multi-dimensional summarization across 9 domains, addressing gaps in existing benchmarks.

Yuho Lee, Taewon Yun, Jason Cai et al.

Proposes a domain-adapted reward model for offline ad ranking evaluation, outperforming IPS and non-generalized models.

Mohamed A. Radwan, Himaghna Bhattacharjee, Quinn Lanners et al.

Human-centered quadruped guide dog robot utilizing visual foundation models and RL control, achieving 95% obstacle avoidance success and 2-hour battery life.

Hochul Hwang, Ken Suzuki, Nicholas A Giudice et al.