NOFACE3D is a training‑free framework that refines fine‑scale, detail‑critical regions in 3D assets generated from a single image. It extracts a selected submesh, scales it to the generator’s input volume, and re‑encodes it into a full‑resolution latent grid, refining the region with flow‑based sampling guided by a spatially varying noise schedule. The refined region is then reintegrated, preserving surrounding geometry and appearance, and the method is evaluated on PAniC‑3D, TexVerse, and a new BigDetail benchmark, showing sharper geometry and finer texture detail.
whyItMatters: The approach improves the fidelity of small, high‑curvature structures in image‑to‑3D models without requiring additional training, addressing a key limitation in current generators.
By James Rose, Jingwen Liang, Wenchao Ma, Oren Jacob, Haomiao Jiang
FPD-Drive is an end‑to‑end autonomous driving planner that uses future‑privileged multi‑target distillation to transfer diverse, high‑quality trajectory targets from a teacher model with access to future observations to a student policy. It also introduces safety‑aware score correction to improve candidate selection by adjusting rankings between safe and unsafe trajectories. The approach achieves state‑of‑the‑art performance on NAVSIM v1 and v2 benchmarks and demonstrates strong zero‑shot closed‑loop results on HUGSIM.
By Li-Heng Chen, Qifeng Chen, Chengwei Wu, Yingjie Zhang, Sheng Yang, Hongbo Fu, Shaoqing Ren
JMT-YOLO is a new joint multi-task learning framework for domain adaptive object detection that builds on the Mean Teacher architecture. It adds an auxiliary semantic segmentation task and a masking mechanism to better model foreground–background boundaries and improve instance-level alignment. Experiments on Pascal VOC→Clipart1k, Cityscapes→Foggy Cityscapes, and KITTI→Cityscapes show that JMT‑YOLO outperforms existing state‑of‑the‑art methods, achieving 46.9%, 57.3%, and 61.0% detection performance respectively.
By Zhenhua Yu, Jiahao Zhang, Wenchao Zhang, Ou Ye, Hengxiang Zhao
The paper examines how fine‑tuned defenses against prompt injection in large language models can be audited before deployment. By leveraging intermediate checkpoints from the fine‑tuning process, the authors create strong optimization‑based attacks (e.g., GCG and BEAST) that dramatically increase attack success rates—from near zero to up to 96%—against state‑of‑the‑art defenses. This demonstrates that current empirical evaluations often overestimate robustness and that pre‑deployment stress‑testing can uncover real vulnerabilities.
By Xiaoxue Yang, Bozhidar Stevanoski, Matthieu Meeus, Yves-Alexandre de Montjoye
The paper "ABLEIST: Intersectional Disability Bias in LLM-Generated Hiring Scenarios" audits six large language models across 2,820 hiring scenarios that vary by disability, gender, nationality, and caste. It introduces the ABLEIST framework—comprising five ableism-specific and three intersectional harm metrics—to detect subtle biases. The study finds that many state‑of‑the‑art models exhibit significant ABLEIST harms, especially for gender- and caste-marginalized disabled candidates, exposing blind spots in current safety tools.
By Mahika Phutane, Hayoung Jung, Matthew Kim, Tanushree Mitra, Aditya Vashistha
The paper introduces QCLR, a plug‑in module that uses dual‑level quantization to robustify contrastive learning models for skeleton‑based human action recognition. By breaking gradient flow and absorbing small perturbations, QCLR protects representations from adversarial attacks while maintaining quality through an attention‑guided multi‑faceted quantization strategy. Experiments on benchmark datasets show that adding QCLR to existing contrastive learning models significantly improves robustness to both white‑box and black‑box attacks without sacrificing classification accuracy.
By Nanzhu Chen, Feixiang He, Yunfeng Diao, Duygu Sarikaya, He Wang
LETHE is a new protection method that defends against unauthorized image‑to‑video generation by disrupting how diffusion models read a reference image. It achieves this by collapsing the spread of reference value vectors and attracting subsequent frames to these collapsed tokens, thereby severely degrading generated videos. Extensive experiments on modern I2V models and two human‑centric benchmarks show that LETHE outperforms existing protection techniques.
By Boheng Li, Yueming Xing, Yiming Li, Jie Zhang, Tianwei Zhang
SplatQ introduces a combinatorial pruning algorithm for 3D Gaussian Splatting that coordinates regional budget allocation and joint Gaussian selection under a fixed budget. The method partitions Gaussians into spatially and appearance‑similar blocks, smooths block retention quotas to avoid budget concentration, and builds a quadratic approximation of joint deletion effects using pixel color responses. Solving the resulting QUBO with simulated annealing or a coherent Ising machine, SplatQ achieves superior reconstruction quality, spatial balance, and rendering speed across five benchmarks compared to existing pruning techniques.
By Hongbin Ye, Haixu Liu, Shenwei Kang, Wenxiu Wei, Tinglan Wang, Hongdong Zhu, Fan Tang, Qi Gao, Yin Ma, Kai Wen, Hai Wei, Nuo Qian, Wei-Shi Zheng, Shaobo Chen, Fa-Ting Hong, Jingke Meng
The paper introduces a Spatial Context Modeling (SCM) module that traverses feature maps in four orthogonal directions to capture interactions among directional representations, thereby building richer scene layout representations for cross-view geo-localization. It also proposes a High-Frequency Information Guidance (HFIG) module that extracts stable structural cues from the frequency domain to emphasize edges and contours, improving robustness to appearance variations. Experiments show the method performs competitively on standard benchmarks and remains robust under challenging scenarios with unknown orientations and limited fields of view.
By Chao Yang, Songlian Li, Zhigang Tu
The paper presents a thyroid nodule segmentation framework for ultrasound images that combines DeepLabV3+ with EfficientNet‑B5, channel‑attention modules (Squeeze‑and‑Excitation, CBAM, GCA), and a post‑hoc conformal prediction layer to provide pixel‑level uncertainty estimates. Experiments on the TN3K benchmark show that the CBAM‑augmented model with Dice‑BCE loss outperforms existing methods. The approach aims to improve the first stage of a CAD pipeline for malignancy classification and metastatic risk assessment.
By Francesco Teodori, Pierpaolo Brutti, Umberto Ferraro Petrillo
The paper presents a pilot study on generating explanations for satirical news from Australia and the United States, highlighting the difficulty of incorporating recent and local context. It proposes using framing explanations based on unreliable narrator theory with open‑weight models as a cost‑effective alternative to frontier models, and finds that these models more accurately identify satirical tropes in American articles than in Australian ones. The study also introduces a new dataset and evaluation benchmark for satirical explanation.
By Kanay Gupta, David Y. Liu, Aditya Joshi
SOAR (Source-Aware Reinforcement learning) is a new paradigm that transforms monolithic reasoning into source-explicit trajectories for search agents. It disentangles external evidence ($E_{ext}$) and internal knowledge ($E_{int}$) into parallel, observable streams via a Source-Aware Workflow and uses a Source-Consistent Reward to guide policy evolution. Evaluations show that SOAR improves performance on 3B and 7B backbones by 7.3% and 5.4% respectively, while reducing evidence fabrication by 74.8%.
By Shengze Huang, Zihao Tang, Jinluan Yang, Xin Yu, Zhenghao Lin, Xiao Liu, Zengxuan Wen, Yeyun Gong, Haizhen Huang, Weiwei Deng, Feng Sun, Qi Zhang, Fei Wu, Kun Kuang
The paper introduces POrchestra, a proactive agentic orchestration framework that lets sub‑agents request harness adjustments during execution, enhancing adaptability to dynamic environments. It also proposes On‑Policy Orchestration Distillation (OPOD), which uses hindsight feedback from sub‑agents’ requests to provide targeted improvement signals for training the orchestration policy. Experiments on three long‑horizon benchmarks show that POrchestra boosts task success by about 9.1% over the best baseline, and with OPOD the improvement rises to 10.0%, while also improving cost‑performance trade‑offs.
By Xueyang Feng, Chaozhuo Li, Xiaohe Bo, Jiakai Tang, Chenxi Li, Qiwei Ye, Xu Chen
The study evaluates machine‑generated text detectors on scientific literature, finding that many detectors perform near perfectly on certain benchmarks due to superficial formatting cues rather than genuine linguistic differences. A logistic regression using 27 superficial features achieved macro‑F1 scores between 0.833 and 0.997, outperforming other detectors on arXiv subsets. The most powerful cue was the PDF extraction artifact—human texts preserve line breaks while generated texts appear as a single block—allowing classification at 0.996 AUC. By introducing format counterfactuals and normalizing formatting, the authors demonstrate that re‑wrapping machine‑written abstracts can flip 20–27% of detector decisions, reducing the generalization gap and improving out‑of‑domain performance.
By Anastasia Voznyuk, Andrey Grabovoy
The paper "Can machines reason non-verbally?" argues that current AI systems, including DeepSeek, ChatGPT, and Gemini, struggle with visual non-verbal reasoning tasks such as completing abstract patterns or delineating ambiguous regions. It introduces GenVerOpt, a framework that combines generation, verification, and optimisation to enable per-sample reasoning, mirroring human problem-solving. Experiments show GenVerOpt surpasses humans on abstract IQ puzzles, matches clinicians in prostate cancer segmentation, outperforms experts in liver vessel reconstruction, and beats state-of-the-art methods in tumour boundary alignment.
By Shaheer U. Saeed, Zion Tse, Gregory Slabaugh, Daniel C. Alexander, Yipeng Hu
ECGWmark is a fully convolutional, quantization‑free deep watermarking framework designed for electrocardiogram (ECG) signals. It embeds patient metadata into ECGs while preserving high‑frequency diagnostic details by operating in a continuous latent space and using a localized frame‑level mechanism to resist severe temporal desynchronization. A differentiable Physiological Loss guided by a clinical classifier helps maintain diagnostic semantics, achieving over 90% channel‑bit extraction accuracy under circular phase shifts and demonstrating mixed effects on class‑level diagnostic metrics in the MIT‑BIH benchmark.
By Kuan-Yu Chen, Kuan-Chen Wang, Yu Tsao, Jian-Jiun Ding
The paper introduces HF‑WM, an action‑conditioned world model that learns from large‑scale egocentric human videos and incorporates failure dynamics through post‑training. It bridges the embodiment gap with a dual‑branch framework that couples video generation to stabilized pointflow prediction, unifying motion representation across humans and robots while reducing egocentric camera motion. Failure modeling is enhanced by Group Relative Policy Optimization on perturbed actions, guided by action alignment and physical plausibility rewards, enabling the model to anticipate unsuccessful executions. Experiments show HF‑WM outperforms existing methods in action‑conditioned video generation and is useful for evaluating policies in downstream robotic tasks.
By Peiyan Tu, Cong Wang, Hanxin Zhu, Xin Jin, Zhibo Chen
The paper introduces IsoMamba, a Vision Mamba variant that addresses state contention in underwater image enhancement by organizing recurrent states according to token compatibility. It employs a cluster–route–isolate paradigm to route update‑compatible tokens into independent state streams and adds a dynamic channel‑wise reweighting objective to handle wavelength‑dependent reconstruction challenges. Experiments show IsoMamba outperforms existing methods on multiple underwater enhancement benchmarks while producing cleaner hidden state responses.
By Yan Luo, Laibin Chang, Zefan Feng, Yi'ang Chen, Kui Jiang, Zhengguo Li
The paper introduces Query‑Guided Spatial‑Temporal‑Frequency (QSTar) interaction for Audio‑Visual Question Answering, integrating question cues and audio frequency characteristics with spatial and temporal visual features. It also presents a Query Context Reasoning (QCR) block that prompts the model to focus on semantically relevant audio and visual elements. Experiments on two AVQA benchmarks show significant performance gains over existing Audio QA, Visual QA, Video QA, and AVQA methods.
By Kun Li, Michael Ying Yang, Sami Sebastian Brandt
The paper introduces Multi-Modal Typography, a systematic investigation into how perturbations applied to speech and on-screen text can compromise audio‑visual reasoning in large language models. Experiments on the WorldSense benchmark show that such attacks can raise the attack success rate against Gemini 3.7 Flash from 9.53% to 43.41%, and fine‑tuning can reduce the success rate for Qwen2.5‑Omni‑7B from 60.6% to 52.0%. The study highlights that cross‑modal typography attacks are a significant, yet underexplored, vulnerability in multi‑modal reasoning systems.
By Tianle Chen, Deepti Ghadiyaram