Past ArXiv / August 14, 2026

Daily archive

A preserved paper digest from the previous month.

Archive date 08/14/2026

Source ArXiv

Back to current digest

Reading Queue

ArXiv

18 papers

Vision-Language Reasoning

1 LongEarth-R1: Benchmarking and Aligning Vision-Language Models for Long-Horizon Earth Observation Reasoning Yupan Ding, Jing Xiao, Zhenyuan Zhang, Chaofeng Chen, Liang Liao, Gui-Song Xia, Mi Wang

Vision-Language Reasoning Benchmark & Evaluation Time-Series LLM

ArXiv

Paper 1 / arXiv:2608.13344 Open arXiv

Relevance 8 Novelty 7

Why selected: Matches criterion 2 and 4 very closely: it benchmarks and aligns vision-language models for long-horizon Earth observation reasoning, with structured reasoning traces and reward-based alignment.

Long-horizon Earth observation reasoning requires models to organize multi-stage geographic evolution, localize spatial changes, detect temporal anomalies, and infer future from extended image sequences. However, existing remote sensing vision-language models mainly focus on isolated images, image pairs, or short sequences, limiting reliable grounding in the relevant frames and regions. We introduce LongEarth-Bench, a benchmark containing approximately 120k question-answering samples derived from 117k unique images. Its sequences average 15.14 frames and extend to 30 frames, covering 12 tasks across evolution summarization, spatial reasoning, anomaly identification, and logical prediction. A 30k-sample subset further provides structured reasoning traces linking key frames and changed regions to final answers. We develop LongEarth through supervised fine-tuning with explicit sequence identifiers and structured chain-of-thought supervision. Building on LongEarth, LongEarth-R1 applies group relative policy optimization with format, temporal, and spatial rewards. LongEarth-R1 achieves the best results on all 12 long-sequence tasks while remaining competitive on standard remote sensing benchmarks.

7 OmniScientist: An Omni-Modal Omni-Discipline AI Scientist Bobo Li, Hao Fei, Tianjie Ju, Mong-Li Lee, Wynne Hsu

Vision-Language Reasoning Benchmark & Evaluation Embodied Interaction Generative Modeling

ArXiv

Paper 7 / arXiv:2608.13558 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 4: a foundation-model-based omni-modal AI scientist that reasons over raw multimodal evidence, including images, video, 3D structures, trajectories, and graphs.

Recent advances in foundation models have enabled AI scientists to automate increasingly complete research workflows, from hypothesis generation and code execution to manuscript preparation. Yet workflow coverage alone does not provide access to the full evidence on which scientific discovery depends. Existing systems typically reason over text, code, labels, or precomputed summaries, leaving scientifically decisive spatial, temporal, cross-channel, and procedural relations unavailable to the agent. We introduce OmniScientist, an end-to-end, omni-modal AI scientist that conducts multidisciplinary research directly from heterogeneous raw evidence. A perception layer and 3 autonomous agents for ideation, experiment, and writeup operate within a deterministic pipeline, allowing observations to shape research questions, experimental decisions, and final claims throughout the research lifecycle. By running idea, rigour, and claim checks in code, the system enforces novelty screening, statistical validity, execution provenance, and numerical traceability. We evaluate OmniScientist on 36 real-data cases spanning 5 discipline families, 4 families of scientific evidence, and modalities including images, signals, audio, video, 3-D structures, trajectories, tables, formulae, and graphs. The system completes the full path from raw data to a compiled manuscript in all 36 cases and achieves a mean overall paper score of 6.3 with the reference reasoning backbone. In paired comparisons against a blind variant that receives only precomputed scalar features, direct perception improves all 7 evaluation dimensions and wins 85% of head-to-head judgments. These results show that lifecycle-wide perception is essential for evidence-grounded scientific discovery and provides a practical path toward broadly capable AI scientists.

9 GS$^{2}$CI: Robust Gaussian Splatting For Snapshot Compressive Imaging via Large Vision Model Priors Yanming Yang, Chenxi Song, Ping Wang, Xin Yuan, Chi Zhang

Vision-Language Reasoning Benchmark & Evaluation Time-Series LLM

ArXiv

Paper 9 / arXiv:2608.13502 Open arXiv

Relevance 6 Novelty 5

Why selected: Matches criterion 4: it uses large vision foundation model priors for 3D reconstruction from snapshot compressive imaging, which is a strong vision-foundation-model application.

Snapshot Compressive Imaging (SCI) offers an efficient solution for high-speed video acquisition and, under exposure-time camera--scene relative motion, multi-view scene capture by compressing temporal or spatial information into a single 2D measurement. While recent studies have explored SCI for 3D scene reconstruction, existing methods struggle with significant challenges due to information loss, limited viewpoint diversity, and the computational burden of jointly optimizing 3D representations and camera poses. In this work, we propose a novel framework that reconstructs high-quality 3D scenes from a single SCI measurement by leveraging 3D Gaussian Splatting (3DGS) and the powerful priors of large-scale vision foundation models (VFMs). Our primary reconstruction combines measurement-derived 3D VFM initialization with SCI-aware Gaussian optimization. After coarse-stage convergence, an auxiliary 2D VFM provides pseudo-view supervision at synthesized viewpoints for local appearance refinement. To further address the instability caused by ambiguous SCI supervision during 3DGS optimization, we introduce Opacity-Guided Splitting and Growth Regulation (OSGR), an SCI-specific densification strategy that augments split candidates using local opacity statistics, discourages loss-compensating opacity inflation through mean-opacity regulation, and bounds representation growth with explicit candidate-ratio and Gaussian-count constraints. Extensive experiments across multiple benchmarks demonstrate that our method achieves the strongest overall performance, combining leading reconstruction quality and robustness to viewpoint variation with competitive computational efficiency.

17 P2Fusion: Prompt-based Progressive Infrared-Visible Image Fusion via Dual-Prior Distillation Yi Shi, Huichao Xie, Yuqing Wang, Mingyu Wang, Kaihui Yang, Yu Liu, Ruitao Lu, Lizhe Li, Junwei Han, Dingwen Zhang

Vision-Language Reasoning Multimodal LLM Benchmark & Evaluation Safety & Backdoors

ArXiv

Paper 17 / arXiv:2608.13045 Open arXiv

Relevance 3 Novelty 4

Why selected: No strong match to the criteria; this is an infrared-visible image fusion method, which is relevant to computer vision but not specifically to the requested VLLM/embodied/benchmark/foundation-model themes.

Infrared-visible image fusion (IVIF) is pivotal for multimodal perception, yet reconciling the inherent information disparity between thermal and textural features remains a fundamental challenge. Existing prior-guided methods often rely on static constraints that induce optimization conflicts or utilize extrinsic semantic priors from large-scale foundation models (e.g., CLIP/DINO), which frequently fail to exploit the intrinsic modality characteristics essential for high-fidelity fusion. To address these issues, we propose P2Fusion, a prior-guided distillation-based framework that reformulates IVIF via dual intrinsic prompts. Instead of imposing hard-coded penalties, we distill image-intrinsic priors, thermal saliency and spatial quality, into learnable dynamic regulators. Specifically, a Teach-to-Fuse mechanism provides dual-granularity progressive guidance, coupled with a Gated Dynamic Expert Recalibration (GDER) module for decoupled feature refinement. This design enables the network to adaptively mediate modal competition through expert specialization. Extensive experiments demonstrate that P2Fusion achieves state-of-the-art performance across five mainstream datasets. Notably, our framework demonstrates consistent performance advantages in fusion quality, achieving state-of-the-art results in 14 out of 20 key evaluation metrics across 5 benchmarks. Furthermore, it effectively contributes to the robustness of downstream perception, such as +3.2% mAP on MSRS, +0.5% mAP on M3FD and +0.9% mAP on DroneVehicle for object detection. Our code will be available at https://github.com/YiShi99/P2Fusion

18 AmalthAI: An Open-Source Computer Vision Platform for Cultural Heritage Christos Chatzisavvas, Stelios Alvanos, Efstratios Politis, Panagiotis Rigas, Thomas Pappas, Ioannis Giannoukos, Nikolaos Mitianoudis, Agata Ulanowska, Katarzyna \.Zebrowska, Nazarij Bu{\l}awka, Christina Margariti, George Pavlidis, Chairi Kiourt, Anestis Koutsoudis, Vassilis Katsouros, George Ioannakis

Vision-Language Reasoning Benchmark & Evaluation Safety & Backdoors LLM Interpretability

ArXiv

Paper 18 / arXiv:2608.13343 Open arXiv

Relevance 4 Novelty 3

Why selected: Matches criterion 4: vision foundation models are used as part of a computer vision platform, but this is more of an application/platform paper than a new foundation model method.

Computer vision (CV) and machine learning (ML) offer new tools for cultural heritage (CH) artifact analysis, but the CV/ML pipeline remains largely inaccessible to CH domain experts, who lack the background to configure, train, or assess models. We present AmalthAI, an open-source CV platform that bridges this gap, enabling non-ML CH experts to independently produce and validate archaeologically meaningful findings. The interface covers dataset management, training, and inference for classification, segmentation, and object detection, with Kubeflow and Katib handling scalable training and hyperparameter search. Grad-CAM localizes the image region behind a prediction, and a vision-language model (VLM) adds a text description of it for expert review. Since archaeological data is often state-owned or rights-encumbered and cannot leave institutional custody, AmalthAI's self-hostable deployment ensures sensitive data is kept within premises. We test the platform on an archaeological use case built on a custom dataset of clay textile imprints, where CH experts trained and validated segmentation, and classification models for hypothesis testing. We provide the implementation code at https://github.com/TEXTaiLES/AmalthAI.

Multimodal LLM

2 TennisVAR: A Stroke-Evidence-Grounded Multimodal Large Language Model for Tactical Reasoning in Tennis Videos Yifan Mei, Qingling Shi, Changli Wu, Jiayuan Rao, Jiayi Ji, Liujuan Cao

Multimodal LLM Benchmark & Evaluation Time-Series LLM Retrieval & Grounding

ArXiv

Paper 2 / arXiv:2608.12920 Open arXiv

Relevance 9 Novelty 6

Why selected: Matches criterion 2 and 3 very closely: it is a new MLLM for tactical reasoning in tennis videos, and it also introduces a new benchmark with grounded evidence annotations.

Sports-video understanding is moving beyond event recognition toward explaining how actions collectively shape match progression, however, existing tennis-video methods either perceive individual strokes without modeling their tactical dependencies or generate high-level analyses without grounding them in the underlying events. To bridge this perception-to-understanding gap, we formulate stroke-evidence-grounded tactical reasoning, a new rally-level task that requires models to jointly predict an open-ended answer, a hierarchical tactic label, an ordered sequence of supporting strokes, and decisive key actions, with each evidence stroke anchored to its racket-ball contact frame. We further introduce TRACE (Tactical Reasoning with Action-Chain Evidence in Tennis), a large-scale expert-annotated benchmark containing 11,189 rally videos, 41,485 stroke events, 25,429 tactical units, and 11,189 question-answer pairs, which unifies fine-grained stroke attributes, cross-stroke tactical relations, hierarchical tactic annotations, and evidence-grounded questions across factual perception, tactical understanding, and decision reasoning. Building on TRACE, we propose TennisVAR (Tennis Video Action-chain Reasoner), an evidence-grounded multimodal large language model that follows an "event-relation-evidence-tactic" reasoning paradigm, where an Event Parsing Module converts continuous rallies into explicit stroke-event sequences while a Tactical Graph-Guided Temporal Reasoner jointly models rally progression and same-player decision dependencies to identify question-relevant evidence and decisive actions.

3 HounsWorld: A Multimodal World Model for Hidden Patient-State Readout, Reconstruction, and Simulation Yunhao Bai, Zhongwei Qiu, Guangyu Guo, Yiming Huang, Tony C. W. Mok, Qinji Yu, Ling Zhang, Yan Wang

Multimodal LLM Benchmark & Evaluation Medical AI Generative Modeling

ArXiv

Paper 3 / arXiv:2608.12904 Open arXiv

Relevance 6 Novelty 7

Why selected: Matches criterion 2 very closely: it introduces a new multimodal world model (HounsWorld) for CT+language understanding/generation, and also fits criterion 4 via a medical vision foundation model application.

Clinical intelligence requires estimating a patient's underlying condition from incomplete observations rather than learning isolated mappings from scans to answers. Volumetric medical images provide dense observations of anatomy, attenuation, and lesions, whereas clinical language provides sparse but complementary semantic observations. We formulate CT-centered intelligence as inference over a shared latent patient state, under which readout, reconstruction, and simulation all become state-dependent prediction problems. To operationalize this view, we introduce HounsBench, a computed tomography (CT) centric patient-state benchmark that unifies these three task families with patient-disjoint splits and per-family metrics, and HounsWorld, a 3B multimodal world model that treats volumetric scans and language as observations of the shared state through Joint Understanding-Generation Learning. A shared transformer forms an implicit patient-state estimate and supports three outputs: query-conditioned answers that read out the state, reports and captions that reconstruct it in language, and condition-specific CT volumes for low-dose denoising, virtual contrast enhancement, and anatomy-constrained text-and-mask-to-volume generation. Zero-initialized CT adapters preserve pretrained multimodal mappings, while condition-explicit Hounsfield-unit window sampling exposes clinically meaningful density observations. HounsWorld shows strong performance across all three task families while consistently improving CT understanding through clinically structured completion. Our project is available at https://github.com/byhwhite/HounsWorld.git

4 NARU: A Benchmark for NARrative Evolution and Cultural Nuance Understanding in Japanese Extreme Long Video Yuheng Huang, Jianlang Chen, Jiayang Song, Hua Qi, Aza Kai, Vincent Markert, Edison Marrese-Taylor, Jianjun Zhao, Lei Ma

Multimodal LLM Benchmark & Evaluation Generative Modeling Retrieval & Grounding

ArXiv

Paper 4 / arXiv:2608.13210 Open arXiv

Relevance 7 Novelty 6

Why selected: Matches criterion 2 and 3: it builds a benchmark for long-form video understanding with narrative evolution and cultural nuance, aimed at improving MLLMs on high-context video.

Long-form video understanding encompasses tasks that go beyond retrieving isolated events, including tracking an evolving narrative and interpreting social meaning that may remain implicit. However, existing benchmarks rarely evaluate these capabilities jointly, particularly in high-context, non-English media. To address this gap, we introduce NARU, a benchmark designed to evaluate Narrative evolution and Reasoning on cultural Understanding in Japanese long-form video. NARU consists of 1,481 questions grounded in 155 videos totaling 146.8 hours, spanning four narrative and five cultural dimensions. To construct the benchmark at this scale, we propose a hierarchical memory-based annotation pipeline that transforms raw video into structured event, narrative, and cultural annotations, then generates questions via task-oriented synthesis and iterative shortcut removal. The construction process includes two native-speaker verification stages involving 68 annotators. Evaluations across eight model configurations reveal substantial limitations in both long-range narrative integration and culturally grounded reasoning. By exposing these persistent gaps, NARU offers a systematic testing ground for developing MLLMs capable of reliably interpreting long-form, high-context video.

5 Towards Physics-Faithful Generation of Scientific Diagrams Minghui Zhang, Jinxin Shi, Yifan Chang, Liangliang Zhao, Yuandong Pu, Qian Yu, Ming Hu, Hanxiao Zhang, Yun Gu, Yirong Chen, Yu Qiao, Bo Zhang, Xiangchao Yan, Bin Fu, Yihao Liu

Multimodal LLM Generative Modeling LLM Interpretability Retrieval & Grounding

ArXiv

Paper 5 / arXiv:2608.13112 Open arXiv

Relevance 7 Novelty 6

Why selected: Matches criterion 4: this is a generative multimodal system for physics-faithful scientific diagram generation, leveraging structured reasoning and multimodal supervision.

Text-to-image generation has reached photorealistic quality, yet state-of-the-art systems remain unreliable at producing scientific diagrams, whose value depends not on appearance but on physical faithfulness: correct force directions, valid coordinate systems, consistent thermodynamic states, and equations matching the depicted scenario. Trained on web imagery with physically shallow captions, generic models produce diagrams that look plausible but are physically wrong, harmful in education and scientific communication. We present Princigram, a physics-faithful scientific-diagram generator, and its data pipeline. Our central advance is Structured Physical Chain-of-Thought (SP-CoT): a per-subdiscipline schema that decomposes a physics diagram into an explicit multi-step reasoning chain across six subdisciplines, from scene identification through force or process analysis to governing laws and synthesis. Unlike free-form chain-of-thought, SP-CoT follows a fixed schema with strict fidelity rules that separate visually grounded facts from physically inferred reasoning and type all mathematics symbolically; it serves both as dense training supervision and, at inference, as a structured "thinking" prompt. With it we curate and structurally annotate 4.3 million physics images, of which 115,037 carry expert-level annotation, and adapt a unified multimodal backbone. We further introduce VeriphyT2IBench, whose questions are derived from each held-out diagram's own structured annotation: each diagram becomes an item-specific bank of binary questions about its objects, forces, and states, so a judge model's score decomposes into named physical facts rather than one holistic number. On the physics subset of GenExam and on VeriphyT2IBench, Princigram shows that explicit physics-structured supervision improves the physical faithfulness of generated scientific diagrams.

8 EgoCITE: Context-Augmented Indexing and Time-Aware Retrieval for Long-Horizon Egocentric Memory Le Zhang, Ke Sun

Multimodal LLM Embodied Interaction Time-Series LLM Retrieval & Grounding

ArXiv

Paper 8 / arXiv:2608.12627 Open arXiv

Relevance 6 Novelty 5

Why selected: Matches criterion 3 in a weaker sense: it introduces a new benchmark and retrieval method for long-horizon egocentric memory, which is closely related to embodied/agentic AI.

Long-horizon egocentric memory transforms continuous first-person video and audio into a searchable record of past experiences. We demonstrate two bottlenecks in existing systems: indices built from context-poor captions are unreliable for agentic search, while retrieval ignores a question's temporal intent. To address both bottlenecks, we introduce EgoCITE (Egocentric Context-augmented Indexing and Time-aware Evidence retrieval), a long-horizon agentic memory framework for egocentric QA. EgoCITE comprises three components. EgoScheme uses local multimodal context to turn fragmentary video captions and speech transcripts into self-contained atomic memory indices. EgoIndex organizes complementary action, activity, utterance, and conversation representations into searchable multi-view memory indices at multiple granularities. EgoRetrv combines semantic search with question-conditioned temporal relevance scoring and curation of retrieved evidence. We evaluate EgoCITE on EgoLifeQA, EgoMem, and EgoR1-Bench in terms of answer accuracy and target-event retrieval alignment. EgoCITE improves accuracy over agentic memory baselines by at least 4.4--14.2\% while achieving 36$\times$ lower cost than long-context LLM agents.

12 UniTraffic-Agent: Unified Traffic Video Reasoning for AI City Challenge 2026 Track 3 with Two Out-of-Domain Evaluations Peng Li, Qianqian Xu, Shilong Bao, Yangbangyan Jiang, Qingming Huang

Multimodal LLM Benchmark & Evaluation Embodied Interaction LLM Interpretability

ArXiv

Paper 12 / arXiv:2608.13031 Open arXiv

Relevance 6 Novelty 4

Why selected: Matches criterion 3: an embodied/agent-style benchmark and method for traffic-video reasoning with out-of-domain evaluations, focusing on how events develop and when interactions occur.

Traffic video understanding has become an important problem in intelligent transportation, as road videos provide direct evidence for accidents, violations, and interactions between vehicles and vulnerable road users. A useful system should explain how a traffic event develops, why it happens, and when the relevant interaction occurs, yet this remains difficult for multimodal large language models (MLLMs) because traffic videos contain sparse events and varied viewpoints. We introduce UniTraffic-Agent, the MR-CAS solution for Track~3 of the 10th AI City Challenge, which includes Traffic Anomaly Reasoning (TAR) and two out-of-domain evaluations: FETV for fisheye traffic events and PSI-VQA for pedestrian intention reasoning. UniTraffic-Agent follows an observe--reason--act--verify workflow that samples timestamped visual evidence, reasons over all questions from the same clip in one request, and converts responses through task-specific action adapters. On the official Public leaderboards, MR-CAS ranks 16th on TAR with a score of 0.5780, 2nd on FETV with 0.4884, and 4th on PSI-VQA with 64.4161. The code is available at https://github.com/Roclp/UniTraffic-Agent.

14 TraVEL: Trajectory-Guided Video Embedding Learning for Driving-Video Retrieval Yi-Chung Chen, Philip Jacobson, Tom Lampo, Yiren Lu, Jin Yao, David I. Inouye, Jing Gao, Danhua Guo, Burhan Yaman

Multimodal LLM Benchmark & Evaluation Safety & Backdoors Retrieval & Grounding

ArXiv

Paper 14 / arXiv:2608.13495 Open arXiv

Relevance 3 Novelty 5

Why selected: No very close match to the listed criteria; it is mainly an application of vision-language embeddings to driving-video retrieval, with some embodied/trajectory supervision but not a new embodied-AI benchmark or core spatial intelligence method.

Efficiently retrieving relevant clips from large-scale driving logs is essential for data curation, model development, and safety analysis. Structured and rule-based retrieval systems can explicitly target driving events, but typically require expert-defined rules, auxiliary data, and multi-stage perception pipelines. Multimodal embedding models offer a simpler and more efficient alternative by representing each video with a single searchable vector. However, general-purpose models often rely on shortcuts from static scene context and struggle to distinguish motion-centric events, such as turning left versus right or accelerating versus decelerating. In this work, we study how to adapt a general-purpose multimodal embedding model to driving-video retrieval. We first fine-tune Qwen3-VL-Embedding on paired clips and reasoning traces from nuReasoning using an InfoNCE objective. While this stage substantially improves overall retrieval, caption supervision alone remains insufficient for fine-grained motion understanding. We therefore introduce TraVEL (Trajectory-Guided Video Embedding Learning), a motion-aware fine-tuning framework that uses ego-trajectory similarity as a reward within Group Relative Policy Optimization. Trajectories serve only as privileged training supervision; retrieval still operates on single-vector video embeddings without ego poses, expert rules, or auxiliary perception outputs. We further construct a driving-video retrieval benchmark from nuReasoning. Experiments show that TraVEL improves motion-centric retrieval across model scales: relative to SFT, it raises longitudinal and lateral mAP by 9.8 and 4.7 points at 2B, with corresponding gains of 7.2 and 1.5 points at 8B. TraVEL thus combines physically grounded supervision with efficient embedding-based search.

Generative Modeling

6 A Controlled Study of Self-Supervised Image and Video Pretraining under Limited Resources Brun\'o B. Englert, Gijs Dubbelman

Generative Modeling Time-Series LLM

ArXiv

Paper 6 / arXiv:2608.13183 Open arXiv

Relevance 5 Novelty 6

Why selected: Matches criterion 4 closely: it is a controlled study of self-supervised pretraining for visual foundation models, with concrete findings about image/video SSL tradeoffs under limited resources.

Visual foundation models are a cornerstone of image and video understanding but typically require large amounts of data and computation. The current scale required for pretraining visual foundation models may be unsustainable or unnecessary, and significant benefits arise when effective models can be obtained with fewer resources. To better understand how self-supervised learning (SSL) objectives behave under resource constraints, we conduct a controlled study of image and video SSL objectives under matched data, architecture, and compute budgets. We compare contrastive, reconstruction, feature-prediction, and diffusion objectives and evaluate both standalone and jointly trained image-video SSL formulations across a diverse set of image and video understanding tasks. Our results show that DINOv2-style pretraining consistently provides the strongest overall performance under limited resources. Furthermore, combining DINOv2 with video SSL objectives such as VideoMAE substantially improves image classification and segmentation performance, but degrades video tracking and camera-pose estimation performance, revealing an important tradeoff between semantic and geometric representation learning. These findings suggest that combining image and video SSL objectives can be beneficial in resource-limited settings, while highlighting the need for improved methods that better balance semantic, temporal, and geometric supervision.

Benchmark & Evaluation

10 SCULPT: Subtractive Composition for 3D Part Generation Sikuang Li, Chen Yang, Jiemin Fang, Jiazhong Cen, Yuhe Wei, Jichen Pang, Wei Shen, Qi Tian

Benchmark & Evaluation Generative Modeling Safety & Backdoors LLM Interpretability

ArXiv

Paper 10 / arXiv:2608.13541 Open arXiv

Relevance 4 Novelty 6

Why selected: Matches criterion 4 loosely: it is a 3D generative modeling paper for part-aware shape generation, which is relevant to vision foundation model applications but not a vision foundation model itself.

Part-aware 3D generation aims to create digital assets that are coherent as complete objects while exposing structural parts for editing, material assignment, animation, and reuse. Existing methods impose this structure outside the native generation loop: segmentation-based methods partition an already generated shape, while additive methods synthesize parts from predefined layouts, boxes, or tokens and then reconcile them into a whole. The former preserves the generated geometry but fixes the object before part boundaries are determined; the latter exposes part cardinality but often leaves shared boundaries vulnerable to gaps, interpenetrations, and material discontinuities. In this paper, we propose SCULPT, a framework that addresses these challenges through subtractive composition. Given a complete object represented in a structured 3D latent space, SCULPT iteratively applies a joint split predictor to generate one extracted part together with the remaining object. The predictor performs a coupled denoising process conditioned on both the image and the current 3D state, so the extracted part and updated remainder are generated together rather than reconciled after generation. The joint split predictor processes both outputs on the union of their native sparse 3D supports, allowing neighboring supports to overlap rather than imposing a disjoint voxel partition. The rollout ends when the remainder support becomes empty or reaches a fixed safety cap, allowing the number of generated parts to adapt to each object within that bound. Extensive experiments demonstrate state-of-the-art geometry on PartObjaverse while preserving strong complete-object reconstruction after part assembly. Results on four dataset images, one text-to-image-generated input, and one real-world photograph further show fine-grained textured part decomposition beyond the benchmark.

11 MV2: Multi-View Multi-Vehicle Driving Dataset for Novel View Synthesis Sanjay Bhargav Dharavath, Hanvitha Saraswathi Mukkamala, Faizan Farooq Khan, Ioannis Kakogeorgiou, Aditya Arun, C V Jawahar, Zakaria Laskar

Benchmark & Evaluation Generative Modeling

ArXiv

Paper 11 / arXiv:2608.12442 Open arXiv

Relevance 6 Novelty 4

Why selected: Matches criterion 3: it introduces a new simulator-like benchmark/dataset for novel view synthesis in dynamic driving scenes with large viewpoint changes.

Differentiable rendering has advanced novel view synthesis (NVS), yet applying it to real-world driving remains difficult due to sparse capture viewpoints, dynamic objects, and limited multi-trajectory data. We introduce the Multi-View Multi-Vehicle (MV2) dataset and benchmark for evaluating NVS models under large viewpoint changes in dynamic urban scenes. MV2 features synchronized captures from a car, scooter, and drone, each following distinct yet synchronized trajectories. Training NVS methods on one vehicle's camera stream and testing on another enables evaluation under substantially larger viewpoint variations than existing single-trajectory datasets. All sequences are registered via Structure-from-Motion and camera poses verified using manual pixel-level correspondence annotations, yielding 50 high-quality scenes with 12000 images. Benchmarking recent NVS and camera pose estimation methods shows that NVS performance degrades with increasing viewpoint disparity, and that feed-forward pose estimators notably lag behind optimization-based approaches, highlighting MV2 as a rigorous testbed for NVS in driving. The dataset, benchmark protocol, and project resources are available at https://mv2-dataset.github.io/.

13 Context-Matched Distillation: Teacher Causality for Autoregressive Video Distillation Hmrishav Bandyopadhyay, Xuanchi Ren, Zijian Huang, Jay Zhangjie Wu, Tianshi Cao, Ruilong Li, Bryan Chu, Sanja Fidler, Yi-Zhe Song, Zian Wang

Benchmark & Evaluation Generative Modeling LLM Interpretability Time-Series LLM

ArXiv

Paper 13 / arXiv:2608.13391 Open arXiv

Relevance 3 Novelty 6

Why selected: No very close match to the listed criteria; it is a methodological video-generation distillation paper, which is adjacent to multimodal generative modeling but not directly about spatial understanding or embodied AI.

Interactive autoregressive video generation demands both low-latency rollouts and precise online control. Few-step distillation accelerates generation by reducing denoising steps, while online control imposes a causal constraint: frames and blocks should depend on history and controls available during generation. Existing video distribution matching distillation (DMD) pipelines, however, often supervise causal few-step students using bidirectional teachers that score complete clips. The score for a target can therefore depend on future frames and controls that were unavailable when the student generated it, misaligning teacher supervision with the student's causal information set. We introduce Context-Matched Distillation (CMD), a causal DMD framework that aligns teacher supervision with the information available when each target is generated. CMD replaces bidirectional full-clip scoring with a causal teacher that evaluates each target without access to future frames or controls. The same causal teacher initializes the few-step student, establishing a consistent causal formulation across teacher training, student distillation, and inference. Beyond aligning the temporal information boundary, Prefix Scoring matches supervision to the student's realized rollout context by evaluating each target under the cached student-generated prefix that produced it. Prefix Corruption further stabilizes training by perturbing unreliable prefixes produced early in training while preserving this target-context alignment. With a simple causal formulation, CMD naturally extends to frame-wise and chunk-wise generation, long video distillation, and camera-conditioned distillation. Experiments demonstrate state-of-the-art aggregate performance among autoregressive methods on both short- and long-video benchmarks, together with substantially improved adherence to time-varying camera controls.

15 Beyond Retrieval: Query-Conditioned Reuse of Long-Horizon Agent Trajectories Yifei Li, Heng Wang, Lingling Zhang, Muye Huang, Xinyu Zhang, Jiashuai Liu, Hang Yan, Rongman Xu

Benchmark & Evaluation Embodied Interaction Retrieval & Grounding

ArXiv

Paper 15 / arXiv:2608.12847 Open arXiv

Relevance 3 Novelty 5

Why selected: No close match to the criteria; this is trajectory-memory reuse for web agents, which is agentic but not embodied spatial understanding or a new embodied benchmark.

Retrieval can identify a past trajectory that may matter, yet it does not specify how an acting agent should use that trajectory after users, entities, constraints, or environment state have changed. We identify this post-retrieval reuse step as a distinct bottleneck for long-horizon trajectory memory and formulate an evaluation framework that holds candidate retrieval, target state, model, decoding, and tool budget fixed while varying the support delivered to the agent. We instantiate the framework with query-conditioned reuse (QCR), a deliberately simple target-bound note that records a reusable procedure, bindings to recover, applicability conditions, and verification requirements. QCR serves to test the reuse hypothesis rather than to claim a universally preferred memory format. Across 2,391 target instances in WebArena, WorkArena, and AppWorld, QCR reaches 62.3% average Success, 10.7 points above Full Trajectory, while using 48.9% fewer online tokens. Summary reranking selects a reusable memory for 94.8% of targets, placing end-task Success within 1.8 points of an oracle reusable selector. Analyses by trajectory length and source--target binding shift show that direct trajectory injection loses much of its utility as traces grow longer or source-specific values change, whereas target-bound support preserves a larger share of the measured gain. The resulting framework separates retrieval quality from the problem of turning retrieved experience into safe, useful support for a new task.

16 HIMEC: Directional Change Representation and Fixed-Interface Decoding for Remote Sensing Image Change Captioning Aysha Ashraf (University of Electronic Science, Technology of China), Shaina Ashraf (University of Bonn), Wafaa I. M. Hussin (University of Electronic Science, Technology of China), Ali Haider (University of Electronic Science, Technology of China), Zhi Lu (University of Electronic Science, Technology of China), Zhenming Peng (University of Electronic Science, Technology of China)

Benchmark & Evaluation LLM Interpretability Time-Series LLM

ArXiv

Paper 16 / arXiv:2608.12502 Open arXiv

Relevance 3 Novelty 4

Why selected: No close match to criteria 1-4; this is remote sensing image change captioning, which is a vision-language application but not one of the targeted paper types.

Remote sensing image change captioning (RSICC) converts bitemporal imagery into a sentence describing semantic changes. Most RSICC methods condition caption decoders directly on fused visual features, leaving intermediate change structure and decoder-interface consistency less studied. We present HIMEC, combining Directional Change Representation (DCR) with fixed-interface decoding. DCR separates signed differences into appearance-oriented, disappearance-oriented, and shared-context streams before fusion. A learned-query encoder converts the fused representation into visually conditioned change-query tokens that form the scene decoder's only sample-dependent memory. A training-only auxiliary phrase decoder supplies caption-derived supervision. With a fixed zero input, the scene decoder maintains the same interface during training and inference. Separately, we evaluate a local-to-scene cascade conditioned on teacher-forced local states during training and autoregressive states at inference. On changed LEVIR-CC validation pairs, these states have a mean cosine distance of 0.69. Regime-matched conditioning recovers most of the associated deficit, whereas permuting state correspondence causes no detectable penalty. These findings are limited to the evaluated cascade. In a matched three-seed comparison, HIMEC reaches a Consensus-based Image Description Evaluation (CIDEr) score of $142.81\pm0.60$ on LEVIR-CC, versus $139.51\pm3.40$ for direct fused-feature memory. On SECOND-CC, fixed-zero and regime-matched diagnostic conditioning reach 75.67 and 76.99 CIDEr, respectively, versus 60.77 for the mismatched cascade. The source code will be made publicly available at https://github.com/ayshaashra/HIMEC upon publication.

Archived papers 18

Top score 15

Average score 10.5