Pith. sign in

REVIEW 2 major objections 4 minor 65 references

The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering

T0 review · 2 major / 4 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read The first EgoCross Challenge shows that a nearly training-free, domain-wise inference strategy can outperform full supervised fine-tuning on cross-domain egocentric video question answering, reaching 66.98% accuracy versus 46.08%.

desk verdict A solid challenge report whose headline result is secure; the fine-grained leaderboard ranks are statistically fragile and should be read with caution. read the letter →

arxiv 2608.04589 v1 pith:E2GUA33U submitted 2026-08-05 cs.CV cs.AI

Yuqian Fu , Tianwen Qian , Yanjun Li , Yu Li , Kunyu Peng , Xu Zheng , Yongqin Xian , Alessio Tonioni
show 25 more authors
Yanwei Fu Xiaoling Wang Danda Paudel Federico Tombari Luc Van Gool Leyi Wu Yifan Zhao Jinjie Zhang Yinchuan Li Yingcong Chen Zixu Li Zhiwei Chen Zhiheng Fu Wenbo Wang Yupeng Hu Weili Guan Liqiang Nie Takuya Murakawa Toru Tamaki Yi Wen Zhenglin Du Zhengyang Li Lingling Li Licheng Jiao Wenping Ma
This is my paper
classification cs.CVcs.AI
keywords egocentricvideoquestionansweringcross-domaingeneralizationmultimodallargelanguagemodelsmultiple-choicebenchmarkdomain-wiseinferenceadaptation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This report presents the first public challenge on EgoCross, a benchmark asking multimodal models to answer multiple-choice questions about first-person videos from four out-of-the-ordinary domains: surgery, industrial assembly, extreme sports, and animal-mounted cameras. The central result is that DomainWiseInfer, a nearly training-free strategy that routes each question to a domain-specific inference procedure, took first place in both the source-limited and open-source tracks with 66.98% overall accuracy, improving on the official supervised fine-tuned baseline by 20.90 percentage points. The authors argue this shows that careful inference-time design—packaging visual evidence, cueing domain knowledge, and deterministically extracting answers—can matter more than additional training when adapting to unfamiliar egocentric domains. A sympathetic reader would take away that cross-domain egocentric video QA is not just a data-scaling problem; the interface between the model and the question is itself a major source of generalization.

What carries the argument

The machinery is threefold: the EgoCross benchmark and its small support set, the two-track competition design, and DomainWiseInfer's domain-wise inference framework. EgoCross is a cross-domain egocentric video QA benchmark with 798 clips and 957 multiple-choice questions across surgery, industry, extreme sports, and animal perspectives, providing the testbed and evaluation protocol. DomainWiseInfer is the nearly training-free framework that carries the argument: it routes each sample according to target domain and question type, then applies tailored media packaging, domain cueing, and deterministic answer extraction—for example, chronological frame labeling and interaction grounding for animal videos, and visibility checks and temporal localization for surgery. This routing decomposes hard questions into verification and grounding subproblems, which is what the paper credits for the 20.90-point gain over the single-pipeline supervised fine-tuned baseline.

What would settle it

Re-run DomainWiseInfer and the supervised fine-tuned baseline on a fresh, unseen split of EgoCross questions drawn from the same four domains. If the method's accuracy falls toward the baseline or the ranking reverses, the claimed inference-strategy advantage is likely an artifact of test-set memorization or probing rather than a true generalization gain.

Watch

Extended reading notes

Core claim

Within the challenge's CloseQA setting (four options, one correct), the paper's central claim is that the winning solution outperforms supervised fine-tuning under strict data constraints. DomainWiseInfer keeps the fixed 4-billion-parameter vision-language backbone largely unchanged—it uses the zero-shot base model for surgery and animal videos and the official two-epoch SFT checkpoint for industry and extreme sports—and obtains 66.98% accuracy in both the Source-Limited and Open-Source Tracks, versus 46.08% for the supervised fine-tuned baseline. The same pattern appears across the leaderboard: animal-perspective questions are answered best (up to 77.05% for the winner), while surgery and extreme sports remain hardest, and the top teams in the Open-Source Track finish within about one percentage point of each other. The authors read these results as evidence that strong average performance does not necessarily mean uniform cross-domain generalization and that test-time reasoning structure is a key ingredient.

Load-bearing premise

The whole comparison rests on the assumption that the 957 test questions are correctly annotated, leakage-free, and that the small accuracy differences separating top teams are not an artifact of more than 1,500 submissions probing the same public test set.

Editorial extensions

If this is right

  • If the reported leaderboard holds, a 4-billion-parameter base model plus an 80-sample support set is sufficient to reach 67% accuracy on four unfamiliar egocentric domains, so the practical bottleneck for such tasks is inference-interface design rather than model scale or training data.
  • The per-domain routing principle implies that future cross-domain egocentric VQA systems will separate questions by capability and domain at test time, and a single universal prompt will be a weak default.
  • The small gap between the top three Open-Source Track entries suggests the field converges on structured reasoning, and further gains will likely come from better temporal evidence normalization and option verification rather than from stronger base models alone.
  • The same method winning both tracks, under different data allowances, indicates the advantage is not driven by additional training data but by the inference protocol itself.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A natural extension beyond the paper: because DomainWiseInfer is nearly training-free, its dominance implies that much of the apparent difficulty of cross-domain egocentric VQA may be an artifact of question formatting and evidence presentation; if so, similar gains should appear on other egocentric benchmarks when the same routing is applied, which is testable.
  • The public test set received over 1,500 submissions, so top scores may partly reflect leaderboard probing; a held-out post-competition evaluation would clarify whether the 66.98% accuracy is robust to unobserved test questions.
  • The winning strategy assumes the test domains are known in advance; a natural extension would evaluate routing when domain labels are unavailable at test time, where the method's advantage could shrink.
  • Animal-perspective questions consistently scoring highest suggests appearance-based questions dominate that domain; a question-type breakdown would reveal whether the per-domain gaps reflect genuine reasoning difficulty or annotation bias.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 4 minor

Summary. This paper reports the organization and results of the First EgoCross Challenge at EgoVis 2026, a cross-domain egocentric video question answering competition. The challenge uses the EgoCross benchmark, which contains 798 video clips and 957 four-choice QA pairs spanning surgery, industrial assembly, extreme sports, and animal perspectives. Two tracks are described: Source-Limited, which restricts participants to the Qwen3-VL-4B baseline and an 80-sample support set, and Open-Source, which allows broader models and data while prohibiting manual construction of target-domain training data. The paper presents the official baseline (zero-shot and SFT), top-10 leaderboard tables for both tracks, and summaries of the four award-winning solutions (DomainWiseInfer, OmniEgo-R2, Reflective Dialogue, TokenInj-RAA). The central reported result is that DomainWiseInfer achieved 66.98% overall accuracy in both tracks, improving on the SFT-Qwen3VL baseline by 20.90 percentage points.

Significance. If the leaderboard is accurate, the challenge provides a useful public resource for studying cross-domain egocentric VQA and demonstrates a striking result: a nearly training-free domain-wise inference strategy outperforms full supervised fine-tuning on a fixed backbone by a large margin. The paper also offers concrete descriptions of the top methods, which may inspire further work on inference-time adaptation, temporal grounding, and structured reasoning for egocentric video. The public release of challenge data, baseline code, and winning-team code is a clear strength that supports reproducibility and follow-up research.

major comments (2)
  1. [Section 2.3 and Section 3] The paper does not state whether the ground-truth answers for the 957 test questions were withheld from participants during the challenge. Since the test set is taken from the public EgoCross benchmark [37], the authors must clarify the annotation secrecy protocol (e.g., whether answers were hidden on Codabench, whether the test set was released with answers, and what measures prevented participants from accessing ground-truth labels). Without an explicit leakage-prevention statement, the validity of all reported leaderboard accuracies is in question.
  2. [Table 2, Section 3] DomainWiseInfer's per-domain accuracies are identical across the Source-Limited and Open-Source Tracks (Surgery 65.72, Industry 64.49, XSports 63.41, Animal 77.05). The paper should state explicitly whether the same prediction file was submitted to both tracks and, if so, whether the Open-Source entry used the same model and inference procedure as the Source-Limited entry. If the submissions were different, the identical per-domain numbers are statistically implausible and require explanation. This clarification is necessary for interpreting the claim of winning both tracks and for understanding what the Open-Source Track actually evaluated.
minor comments (4)
  1. [Section 3] The 0.21 percentage point gap between DomainWiseInfer (66.98%) and OmniEgo-R2 (66.77%) in the Open-Source Track corresponds to roughly 2 questions out of 957. The report should add a caveat that rank differences at this scale are within sampling noise, since the paper provides no confidence intervals or multiple-run analysis.
  2. [Section 1] The paper reports 'more than 1,500 submissions' but does not disclose how many submissions each top team made or whether the reported scores are the maximum over those submissions. A best-of-N selection procedure can inflate leaderboard scores; the authors should clarify the evaluation protocol (e.g., whether the final ranking uses the best or last submission).
  3. [Introduction, first paragraph] The reference list contains a duplicated entry: the citation group '[8, 29–35, 35, 36]' includes '35' twice. Please correct this typo.
  4. [Section 2.2, Eq. (1)] The definition of accuracy uses the notation '⊮[ˆyi = y i]' but the indicator function symbol appears to be a nonstandard glyph. Consider using the standard '1[·]' or 'I(·)' notation for clarity.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the challenge leaderboard is an independent measurement on a fixed test set, and no derived result reduces to its inputs.

full rationale

This paper is a challenge report rather than a derivation. The central claims, namely the rankings and accuracies in Table 2, are computed by automatically matching submitted answers against the fixed EgoCross test labels using Eq. (1); no parameter is fitted to those labels, and no equation defines its output in terms of an input that the paper itself supplies. Reporting that 'DomainWiseInfer ranked first with an overall accuracy of 66.98%' is a measurement of externally submitted inference pipelines, not a construction from the test answers. The benchmark is cited to prior work [37] whose authors overlap with the organizers, but that citation supplies the dataset and labels rather than an unverified uniqueness result or an ansatz; the winning solutions are separate artifacts, several authored by participants outside the organizer list. None of the enumerated circularity patterns (self-definition, fitted-input-as-prediction, load-bearing self-citation, uniqueness import, ansatz-via-citation, or renaming a known result) occurs. The narrow margin between the top Open-Source teams is a statistical-significance concern, but it is not a circularity concern.

Assumptions & free parameters 0 free parameters · 3 assumptions · 1 invented entities

The central claim rests on the EgoCross test set's validity and on the challenge rules being followed. No free parameters are fitted in this report; the baseline hyperparameters (2 epochs, LR 1e-5) are fixed design choices, not fitted to the leaderboard. The support set and benchmark are released publicly, which lowers reproducibility burden.

assumptions (3)
  • domain assumption The 957 multiple-choice QA pairs in the EgoCross test set are correctly annotated and a valid measure of VQA capability.
    The entire leaderboard interpretation rests on the benchmark's correctness; the paper cites [37] for the dataset but provides no independent verification.
  • domain assumption Accuracy on four-way multiple choice, with a 25% random baseline, captures cross-domain generalization.
    Section 2.2 defines the metric and treats it as the evaluation protocol; this is a modeling choice.
  • domain assumption The challenge rules, especially the prohibition of manually constructing target-domain training data in the Open-Source Track, were honored and enforceable.
    Section 2.1 describes the rules; the validity of the open-source comparison depends on compliance.
invented entities (1)
  • EgoCross support set independent evidence
    purpose: A fixed set of 80 QA samples used for model adaptation in the Source-Limited Track.
    Publicly released with the baseline code; however, its small size and selection may bias adaptation comparisons.

how reviews work

0 comments
Cite this review

Pith. "Pith review of The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering." pith.science (2026). https://pith.science/paper/E2GUA33U

@misc{pith2026260804589,
  author       = {Pith},
  title        = {Pith review of: The First EgoCross Challenge at EgoVis 2026: Cross-Domain Egocentric Video Question Answering},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E2GUA33U}},
  note         = {Machine review of arXiv:2608.04589}
}
read the original abstract

EgoCross is a cross-domain egocentric video question answering benchmark designed to evaluate whether multimodal large language models can generalize beyond common daily-life scenarios. The first EgoCross Challenge was hosted at the Third EgoVis Workshop at CVPR 2026 and evaluated models on first-person videos from four target domains: surgery, industrial assembly, extreme sports, and animal perspectives. Each test example consists of an egocentric video clip, a question, and four candidate answers, from which the model must select the correct option. This technical report introduces the challenge task, benchmark resources, and two official Codabench tracks. The Source-Limited Track restricts participants to the official baseline model and a small support set, whereas the Open-Source Track permits broader choices of models and training data under rules that prohibit the manual construction of target-domain training data. In total, the challenge received more than 1,500 submissions from over 130 participants, with 19 teams participating in the Open-Source Track and 38 teams in the Source-Limited Track. We further present the official leaderboard results and summarize the winning solutions from both tracks. We hope that this report will serve as a useful technical reference for advancing cross-domain egocentric video understanding. All resources, including the challenge data, baseline implementation, and code released by the winning teams, are made publicly available.

Figures

Figures reproduced from arXiv: 2608.04589 by the authors.

Figure 1
Figure 1. EgoCross task examples. The challenge covers professional and non-daily egocentric domains, including surgery, industry, extreme sports, and animal perspectives, with each question categorized by its corresponding task type. EgoCross Challenge1 was formally organized as part of the Third EgoVis Workshop at CVPR 20262 to increase its visibility within the community and stimulate further research on cross-domain egoce… view at source ↗
Figure 2
Figure 2. Overview of the two EgoCross challenge tracks. The Source-Limited Track restricts participants to the official Qwen3VL-4B backbone and a small target-domain support set, enabling controlled comparison under limited data and model re￾sources. The Open-Source Track allows participants to use al￾ternative models and additional public data, while prohibiting the manual collection of target-aligned training data. Both tr… view at source ↗
Figure 3
Figure 3. Overview of DomainWiseInfer. The method uses a nearly training-free, domain-wise inference framework with tai￾lored input packaging, domain cueing, and answer extraction. 4. Methods This section summarizes the solutions from the award￾winning teams. Rather than reproducing the full technical details of the original reports, we focus on the key designs most relevant to the EgoCross Challenge, including tempo￾ral grou… view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Overview of OmniEgo-R2 [62]. The framework transforms each egocentric VQA sample into a unified routed reasoning pipeline consisting of Temporal Evidence Normalization, Capability-Oriented Routing, Role-Decomposed Reasoning, and Boundary-Aware Option Verification. Doma…
Figure 5
Figure 5. Figure 5: Overview of Reflective Dialogue [64]. Support examples are grouped by domain and question type and converted into multi￾turn conversations between a Teacher and a Solver. These dialogues contain correctness feedback, visual grounding explanations, and contrastive refle…
Figure 6
Figure 6. Figure 6: Overview of TokenInj-RAA. The method integrates timestamp-aware frame encoding, domain-adaptive LoRA rout￾ing, and CholecT50-based retrieval augmentation to improve tem￾poral reasoning and specialized surgical understanding in cross￾domain egocentric VQA. source domain…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

65 extracted references · 46 canonical work pages

  1. [37]

    Egocross: Benchmarking multimodal large language mod- els for cross-domain egocentric video question answering

    Yanjun Li, Yuqian Fu, Tianwen Qian, Qi’Ao Xu, Silong Dai, Danda Pani Paudel, Luc Van Gool, and Xiaoling Wang. Egocross: Benchmarking multimodal large language mod- els for cross-domain egocentric video question answering. InProceedings of the AAAI Conference on Artificial Intelli- gence, 2026. 1, 2, 3

    work page 2026

  2. [1]

    An outlook into the future of egocentric vision: C

    Chiara Plizzari, Gabriele Goletto, Antonino Furnari, Sid- dhant Bansal, Francesco Ragusa, Giovanni Maria Farinella, Dima Damen, and Tatiana Tommasi. An outlook into the future of egocentric vision: C. plizzari et al.International Journal of Computer Vision, 132(11):4880–4936, 2024. 1

    work page 2024

  3. [2]

    Analysis of the hands in egocentric vision: A survey.IEEE transactions on pattern analysis and machine intelligence, 45(6):6846–6866, 2020

    Andrea Bandini and Jos ´e Zariffa. Analysis of the hands in egocentric vision: A survey.IEEE transactions on pattern analysis and machine intelligence, 45(6):6846–6866, 2020

    work page 2020

  4. [3]

    Challenges and trends in egocentric vision: A survey

    Xiang Li, Heqian Qiu, Lanxiao Wang, Hanwen Zhang, Chenghao Qi, Linfeng Han, Huiyu Xiong, and Hongliang Li. Challenges and trends in egocentric vision: A survey. Machine Intelligence Research, 23(1):1–33, 2026

    work page 2026

  5. [4]

    Objectrelator: Enabling cross-view object rela- tion understanding across ego-centric and exo-centric per- spectives.arXiv preprint arXiv:2411.19083, 2024

    Yuqian Fu, Runze Wang, Bin Ren, Guolei Sun, Biao Gong, Yanwei Fu, Danda Pani Paudel, Xuanjing Huang, and Luc Van Gool. Objectrelator: Enabling cross-view object rela- tion understanding across ego-centric and exo-centric per- spectives.arXiv preprint arXiv:2411.19083, 2024

    arXiv 2024

  6. [5]

    V2-sam: Mar- rying sam2 with multi-prompt experts for cross-view object correspondence.arXiv preprint arXiv:2511.20886, 2025

    Jiancheng Pan, Runze Wang, Tianwen Qian, Mohammad Mahdi, Yanwei Fu, Xiangyang Xue, Xiaomeng Huang, Luc Van Gool, Danda Pani Paudel, and Yuqian Fu. V2-sam: Mar- rying sam2 with multi-prompt experts for cross-view object correspondence.arXiv preprint arXiv:2511.20886, 2025

    arXiv 2025

  7. [6]

    Streameqa: Towards streaming video understanding for embodied sce- narios.arXiv preprint arXiv:2512.04451, 2025

    Yifei Wang, Zhenkai Li, Tianwen Qian, Huanran Zheng, Zheng Wang, Yuqian Fu, and Xiaoling Wang. Streameqa: Towards streaming video understanding for embodied sce- narios.arXiv preprint arXiv:2512.04451, 2025. 1

    arXiv 2025

  8. [7]

    Exo2egosyn: Unlocking foundation video generation models for exocentric-to-egocentric video synthesis.arXiv preprint arXiv:2511.20186, 2025

    Mohammad Mahdi, Yuqian Fu, Nedko Savov, Jiancheng Pan, Danda Pani Paudel, and Luc Van Gool. Exo2egosyn: Unlocking foundation video generation models for exocentric-to-egocentric video synthesis.arXiv preprint arXiv:2511.20186, 2025

    arXiv 2025

Show all 65 references
  1. [8]

    Egosound: Benchmarking sound understanding in egocentric videos.arXiv preprint arXiv:2602.14122, 2026

    Bingwen Zhu, Yuqian Fu, Qiaole Dong, Guolei Sun, Tianwen Qian, Yuzheng Wu, Danda Pani Paudel, Xi- angyang Xue, and Yanwei Fu. Egosound: Benchmarking sound understanding in egocentric videos.arXiv preprint arXiv:2602.14122, 2026. 1

    2026 arXiv

  2. [9]

    Affin-space: Learning affine-invariant representations for 3d spatial understanding with MLLMs

    Zhenyu Lu, Liupeng Li, Jinpeng Wang, Haoqian Kang, Manyuan Zhang, Yan Feng, Ke Chen, and Yaowei Wang. Affin-space: Learning affine-invariant representations for 3d spatial understanding with MLLMs. InForty-third Interna- tional Conference on Machine Learning, 2026. 1

    2026

  3. [10]

    Accessible robot control in mixed reality.arXiv e- prints, pages arXiv–2306, 2023

    Ganlin Zhang, Deheng Zhang, Longteng Duan, and Guo Han. Accessible robot control in mixed reality.arXiv e- prints, pages arXiv–2306, 2023. 1

    2023

  4. [11]

    Tom: A development plat- form for wearable intelligent assistants

    Nuwan Janaka, Shengdong Zhao, David Hsu, Sherisse Tan Jing Wen, and Chun Keat Koh. Tom: A development plat- form for wearable intelligent assistants. InCompanion of the 2024 on ACM International Joint Conference on Pervasive and Ubiquitous Computing, pages 837–843, 2024

    2024

  5. [12]

    Snap, segment, deploy: A visual data and detection pipeline for wearable industrial assistants

    Di Wen, Junwei Zheng, Ruiping Liu, Yi Xu, Kunyu Peng, and Rainer Stiefelhagen. Snap, segment, deploy: A visual data and detection pipeline for wearable industrial assistants. In2025 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pages 1270–1276. IEEE, 2025. 1

    2025

  6. [13]

    Embodied large language mod- els enable robots to complete complex tasks in unpredictable environments.Nature Machine Intelligence, 7(4):592–601,

    Ruaridh Mon-Williams, Gen Li, Ran Long, Wenqian Du, and Christopher G Lucas. Embodied large language mod- els enable robots to complete complex tasks in unpredictable environments.Nature Machine Intelligence, 7(4):592–601,

  7. [14]

    Self in space: Benchmark- ing self-awareness and spatial cognition in uav embodied intelligence, 2026

    Zhishan Zou, Guoyan Sun, Zhiwei Wei, Jiancheng Pan, Yujie Li, Mugen Peng, and Wenjia Xu. Self in space: Benchmark- ing self-awareness and spatial cognition in uav embodied intelligence, 2026. URLhttps://arxiv.org/abs/ 2607.12477

    2026 arXiv

  8. [15]

    Clivis: Unleashing cognitive map through linguistic-visual synergy for embodied visual rea- soning.arXiv preprint arXiv:2506.17629, 2025

    Kailing Li, Qi’ao Xu, Tianwen Qian, Yuqian Fu, Yang Jiao, and Xiaoling Wang. Clivis: Unleashing cognitive map through linguistic-visual synergy for embodied visual rea- soning.arXiv preprint arXiv:2506.17629, 2025. 1

    2025 arXiv

  9. [16]

    Bridging the 2d-3d gap: A hierarchical semantic-geometric map for vision lan- guage navigation.arXiv preprint arXiv:2606.00095, 2026

    Kailing Li, Tianwen Qian, Lijin Yang, Yuqian Fu, Jingyu Gong, Xiaoling Wang, and Liang He. Bridging the 2d-3d gap: A hierarchical semantic-geometric map for vision lan- guage navigation.arXiv preprint arXiv:2606.00095, 2026. 1

    2026 arXiv

  10. [17]

    Afford-vla: Action-aligned vi- sual planning via internalized affordance.arXiv preprint arXiv:2605.24203, 2026

    Runze Wang, Yuqian Fu, Yu Li, Tao Lin, Tianwen Qian, Mohamed Elhoseiny, Bo Zhao, Yanwei Fu, Yu-Gang Jiang, and Xiangyang Xue. Afford-vla: Action-aligned vi- sual planning via internalized affordance.arXiv preprint arXiv:2605.24203, 2026. 1

    2026 arXiv

  11. [18]

    La4vla: Learning to act without seeing via language- action pretraining.arXiv preprint arXiv:2606.27295, 2026

    Tao Lin, Yuxin Du, Yiran Mao, Zewei Ye, Yilei Zhong, Bing Cheng, Yiming Wang, Jiting Liu, Yang Tian, Junchi Yan, et al. La4vla: Learning to act without seeing via language- action pretraining.arXiv preprint arXiv:2606.27295, 2026

    2026 arXiv

  12. [19]

    Evo-depth: A lightweight depth- enhanced vision-language-action model.arXiv preprint arXiv:2605.14950, 2026

    Tao Lin, Yuxin Du, Jiting Liu, Nuobei Zhu, Yunhe Li, Yuqian Fu, Yinxinyu Chen, Hongyi Cai, Zewei Ye, Bing Cheng, et al. Evo-depth: A lightweight depth- enhanced vision-language-action model.arXiv preprint arXiv:2605.14950, 2026

    2026 arXiv

  13. [20]

    Focusable monocular depth estimation.arXiv preprint arXiv:2605.11756, 2026

    Yuxin Du, Tao Lin, Zile Zhong, Runting Li, Xiyao Chen, Jiting Liu, Chenglin Liu, Ying-Cong Chen, Yuqian Fu, and Bo Zhao. Focusable monocular depth estimation.arXiv preprint arXiv:2605.11756, 2026

    2026 arXiv

  14. [21]

    Learning precise affordances from egocentric videos for robotic manipulation

    Gen Li, Nikolaos Tsagkas, Jifei Song, Ruaridh Mon- Williams, Sethu Vijayakumar, Kun Shao, and Laura Sevilla- Lara. Learning precise affordances from egocentric videos for robotic manipulation. In2025 IEEE/CVF International Conference on Computer Vision (ICCV), pages 10581– 105...

    2025

  15. [22]

    Oflow: Injecting object-aware temporal flow matching for robust robotic manipulation.arXiv preprint arXiv:2604.17876, 2026

    Kuanning Wang, Ke Fan, Chenhao Qiu, Zeyu Shang- guan, Yuqian Fu, Yanwei Fu, Daniel Seita, and Xi- angyang Xue. Oflow: Injecting object-aware temporal flow matching for robust robotic manipulation.arXiv preprint arXiv:2604.17876, 2026

    2026 arXiv

  16. [23]

    Ocra: Object-centric learning with 3d and tactile priors for human-to-robot action transfer.arXiv preprint arXiv:2603.14401, 2026

    Kuanning Wang, Ke Fan, Yuqian Fu, Siyu Lin, Hu Luo, Daniel Seita, Yanwei Fu, Yu-Gang Jiang, and Xiangyang Xue. Ocra: Object-centric learning with 3d and tactile priors for human-to-robot action transfer.arXiv preprint arXiv:2603.14401, 2026

    2026

  17. [24]

    Evo-0: Vision-language- action model with implicit spatial understanding.arXiv preprint arXiv:2507.00416, 2025

    Tao Lin, Gen Li, Yilei Zhong, Yanwen Zou, Yuxin Du, Jiting Liu, Encheng Gu, and Bo Zhao. Evo-0: Vision-language- action model with implicit spatial understanding.arXiv preprint arXiv:2507.00416, 2025

    2025

  18. [25]

    Evo-1: Lightweight vision-language-action model with preserved semantic alignment.arXiv preprint arXiv:2511.04555, 2025

    Tao Lin, Yilei Zhong, Yuxin Du, Jingjing Zhang, Jiting Liu, Yinxinyu Chen, Encheng Gu, Ziyan Liu, Hongyi Cai, Yan- wen Zou, et al. Evo-1: Lightweight vision-language-action model with preserved semantic alignment.arXiv preprint arXiv:2511.04555, 2025. 1

    2025

  19. [26]

    Mask2iv: Interaction-centric video generation via mask tra- jectories

    Gen Li, Bo Zhao, Jianfei Yang, and Laura Sevilla-Lara. Mask2iv: Interaction-centric video generation via mask tra- jectories. InProceedings of the AAAI Conference on Artifi- cial Intelligence, volume 40, pages 6091–6099, 2026. 1

    2026

  20. [27]

    An overview of augmented reality.Computers, 11 (2):28, 2022

    Fabio Arena, Mario Collotta, Giovanni Pau, and Francesco Termine. An overview of augmented reality.Computers, 11 (2):28, 2022

    2022

  21. [28]

    Revealing the true potential and prospects of augmented reality in education.Smart Learning Environments, 11(1):2, 2024

    Yiannis Koumpouros. Revealing the true potential and prospects of augmented reality in education.Smart Learning Environments, 11(1):2, 2024. 1

    2024

  22. [29]

    Scaling egocentric vision: The epic-kitchens dataset

    Dima Damen, Hazel Doughty, Giovanni Maria Farinella, Sanja Fidler, Antonino Furnari, Evangelos Kazakos, Davide Moltisanti, Jonathan Munro, Toby Perrett, Will Price, et al. Scaling egocentric vision: The epic-kitchens dataset. In Proceedings of the European conference on comput...

    2018

  23. [30]

    Egovqa-an egocentric video question answer- ing benchmark dataset

    Chenyou Fan. Egovqa-an egocentric video question answer- ing benchmark dataset. InProceedings of the IEEE/CVF International Conference on Computer Vision Workshops, 2019

    2019

  24. [31]

    Egotaskqa: Understanding human tasks in egocentric videos

    Baoxiong Jia, Ting Lei, Song-Chun Zhu, and Siyuan Huang. Egotaskqa: Understanding human tasks in egocentric videos. Advances in Neural Information Processing Systems, 2022

    2022

  25. [32]

    Egoschema: A diagnostic benchmark for very long- form video language understanding.Advances in Neural In- formation Processing Systems, 2023

    Karttikeya Mangalam, Raiymbek Akshulakov, and Jitendra Malik. Egoschema: A diagnostic benchmark for very long- form video language understanding.Advances in Neural In- formation Processing Systems, 2023

    2023

  26. [33]

    Egothink: Evalu- ating first-person perspective thinking capability of vision- language models

    Sijie Cheng, Zhicheng Guo, Jingwen Wu, Kechen Fang, Peng Li, Huaping Liu, and Yang Liu. Egothink: Evalu- ating first-person perspective thinking capability of vision- language models. InProceedings of the IEEE/CVF Confer- ence on Computer Vision and Pattern Recognition, 2024

    2024

  27. [34]

    Omnia de egotempo: Benchmarking temporal understanding of multi-modal llms in egocentric videos

    Chiara Plizzari, Alessio Tonioni, Yongqin Xian, Achin Kul- shrestha, and Federico Tombari. Omnia de egotempo: Benchmarking temporal understanding of multi-modal llms in egocentric videos. InProceedings of the Computer Vision and Pattern Recognition Conference, 2025

    2025

  28. [35]

    Tog- bench: Task-oriented spatio-temporal grounding in egocen- tric videos.arXiv preprint arXiv:2512.03666, 2025

    Qi’ao Xu, Tianwen Qian, Yuqian Fu, Kailing Li, Yang Jiao, Jiacheng Zhang, Xiaoling Wang, and Liang He. Tog- bench: Task-oriented spatio-temporal grounding in egocen- tric videos.arXiv preprint arXiv:2512.03666, 2025. 1

    2025 arXiv

  29. [36]

    Egonight: Towards egocen- tric vision understanding at night with a challenging bench- mark

    Deheng Zhang, Yuqian Fu, Runyi Yang, Yang Miao, Tian- wen Qian, Xu Zheng, Guolei Sun, Ajad Chhatkuli, Xuanjing Huang, Yu-Gang Jiang, et al. Egonight: Towards egocen- tric vision understanding at night with a challenging bench- mark. InInternational Conference on Learning Repre...

    2026

  30. [38]

    Linhai Zhuo, Yuqian Fu, Jingjing Chen, Yixin Cao, and Yu- Gang Jiang. Unified view empirical study for large pre- trained model on cross-domain few-shot learning.ACM Transactions on Multimedia Computing, Communications and Applications, 20(9):1–18, 2024. 1

    2024

  31. [39]

    Linhai Zhuo, Zheng Wang, Tianwen Qian, and Yuqian Fu. Segdp: Source-free cross-domain few-shot learning via se- mantic guided diversity prompting.ACM Transactions on Multimedia Computing, Communications and Applications, 22(5):1–20, 2026

    2026

  32. [40]

    Styleadv: Meta style adversarial training for cross-domain few-shot learning

    Yuqian Fu, Yu Xie, Yanwei Fu, and Yu-Gang Jiang. Styleadv: Meta style adversarial training for cross-domain few-shot learning. In2023 IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 24575– 24584. IEEE, 2023

    2023

  33. [41]

    Meta-fdmixup: Cross-domain few-shot learning guided by labeled target data

    Yuqian Fu, Yanwei Fu, and Yu-Gang Jiang. Meta-fdmixup: Cross-domain few-shot learning guided by labeled target data. InProceedings of the 29th ACM international con- ference on multimedia, pages 5326–5334, 2021

    2021

  34. [42]

    A closer look at conditional prompt tuning for vision-language models.International Journal of Computer Vision, 2025

    Ji Zhang, Shihan Wu, Lianli Gao, Jingkuan Song, Nicu Sebe, and Heng Tao Shen. A closer look at conditional prompt tuning for vision-language models.International Journal of Computer Vision, 2025

    2025

  35. [43]

    Learning attention-guided pyramidal features for few-shot fine-grained recognition.Pattern Recognit., 130:108792, 2022

    Hao Tang, Chengcheng Yuan, Zechao Li, and Jinhui Tang. Learning attention-guided pyramidal features for few-shot fine-grained recognition.Pattern Recognit., 130:108792, 2022

    2022

  36. [44]

    Reliable few-shot learning under dual noises.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    Ji Zhang, Jingkuan Song, Lianli Gao, Nicu Sebe, and Heng Tao Shen. Reliable few-shot learning under dual noises.IEEE Transactions on Pattern Analysis and Machine Intelligence, 2025

    2025

  37. [45]

    less is more

    Ji Zhang, Xu Luo, Lianli Gao, Difan Zou, Heng Tao Shen, and Jingkuan Song. From channel bias to feature redun- dancy: Uncovering the “less is more” principle in few-shot learning.IEEE Transactions on Pattern Analysis and Ma- chine Intelligence, 2026

    2026

  38. [46]

    Flatten long-range loss landscapes for cross-domain few-shot learning

    Yixiong Zou, Yicong Liu, Yiman Hu, Yuhua Li, and Ruix- uan Li. Flatten long-range loss landscapes for cross-domain few-shot learning. In2024 IEEE/CVF Conference on Com- puter Vision and Pattern Recognition (CVPR), pages 23575– 23584. IEEE, 2024

    2024

  39. [47]

    Atten- tion temperature matters in vit-based cross-domain few-shot learning.Advances in Neural Information Processing Sys- tems, 37:116332–116354, 2024

    Yixiong Zou, Ran Ma, Yuhua Li, and Ruixuan Li. Atten- tion temperature matters in vit-based cross-domain few-shot learning.Advances in Neural Information Processing Sys- tems, 37:116332–116354, 2024

    2024

  40. [48]

    Prompt-driven simulation with feature perturbation for cross-domain few-shot object detection

    Linhai Zhuo, Junxi Cai, Tianwen Qian, Qingping Zheng, and Yang Liu. Prompt-driven simulation with feature perturbation for cross-domain few-shot object detection. arXiv preprint arXiv:2608.01348, 2026. URLhttps: //arxiv.org/abs/2608.01348. 1

    2026 arXiv

  41. [49]

    Domain-rag: Retrieval-guided compo- sitional image generation for cross-domain few-shot object detection.Advances in Neural Information Processing Sys- tems, 38:147594–147614, 2026

    Yu Li, Xingyu Qiu, Yuqian Fu, Jie Chen, Tianwen Qian, Xu Zheng, Danda Pani Paudel, Yanwei Fu, Xuanjing Huang, Luc V Gool, et al. Domain-rag: Retrieval-guided compo- sitional image generation for cross-domain few-shot object detection.Advances in Neural Information Processing S...

    2026

  42. [50]

    Divide-and-conquer: Confluent triple-flow net- work for RGB-T salient object detection.IEEE Trans

    Hao Tang, Zechao Li, Dong Zhang, Shengfeng He, and Jin- hui Tang. Divide-and-conquer: Confluent triple-flow net- work for RGB-T salient object detection.IEEE Trans. Pat- tern Anal. Mach. Intell., 47(3):1958–1974, 2025

    1958

  43. [51]

    Cross-domain few-shot object detection via enhanced open-set object detector

    Yuqian Fu, Yu Wang, Yixuan Pan, Lian Huai, Xingyu Qiu, Zeyu Shangguan, Tong Liu, Yanwei Fu, Luc Van Gool, and Xingqun Jiang. Cross-domain few-shot object detection via enhanced open-set object detector. InEuropean conference on computer vision, pages 247–264. Springer, 2024

    2024

  44. [52]

    From global to local: Multi-scale out-of-distribution detection.IEEE Transactions on Image Processing, 32:6115–6128, 2023

    Ji Zhang, Lianli Gao, Bingguang Hao, Hao Huang, Jingkuan Song, and Hengtao Shen. From global to local: Multi-scale out-of-distribution detection.IEEE Transactions on Image Processing, 32:6115–6128, 2023

    2023

  45. [53]

    Remedying target-domain astigmatism for cross-domain few-shot object detection.arXiv preprint arXiv:2603.18541,

    Yongwei Jiang, Yixiong Zou, Yuhua Li, and Ruixuan Li. Remedying target-domain astigmatism for cross-domain few-shot object detection.arXiv preprint arXiv:2603.18541,

  46. [54]

    Classes matter: A fine-grained adversarial ap- proach to cross-domain semantic segmentation

    Haoran Wang, Tong Shen, Wei Zhang, Ling-Yu Duan, and Tao Mei. Classes matter: A fine-grained adversarial ap- proach to cross-domain semantic segmentation. InEuropean conference on computer vision, pages 642–659. Springer,

  47. [55]

    Toward fair and accurate cross-domain medical image segmentation: A vlm-driven active domain adaptation paradigm

    Hongqiu Wang, Wu Chen, Xiangde Luo, Zhaohu Xing, Li- hao Liu, Jing Qin, Shaozhi Wu, and Lei Zhu. Toward fair and accurate cross-domain medical image segmentation: A vlm-driven active domain adaptation paradigm. In2025 IEEE/CVF International Conference on Computer Vision (ICCV)...

    2025

  48. [56]

    Adaptive adversarial cross-domain segmen- tation network for high-resolution remote sensing images

    Jianfen Wei, Ping Yang, Chang Wang, Chunxiang Shi, and Renlong Hang. Adaptive adversarial cross-domain segmen- tation network for high-resolution remote sensing images. IEEE Journal of Selected Topics in Applied Earth Observa- tions and Remote Sensing, 19:3464–3476, 2026

    2026

  49. [57]

    Lightweight frequency masker for cross-domain few-shot se- mantic segmentation.Advances in neural information pro- cessing systems, 37:96728–96749, 2024

    Jintao Tong, Yixiong Zou, Yuhua Li, and Ruixuan Li. Lightweight frequency masker for cross-domain few-shot se- mantic segmentation.Advances in neural information pro- cessing systems, 37:96728–96749, 2024

    2024

  50. [58]

    The devil is in low-level features for cross-domain few-shot seg- mentation

    Yuhan Liu, Yixiong Zou, Yuhua Li, and Ruixuan Li. The devil is in low-level features for cross-domain few-shot seg- mentation. In2025 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 4618–4627. IEEE, 2025

    2025

  51. [59]

    Self-disentanglement and re-composition for cross-domain few-shot segmentation.arXiv preprint arXiv:2506.02677, 2025

    Jintao Tong, Yixiong Zou, Guangyao Chen, Yuhua Li, and Ruixuan Li. Self-disentanglement and re-composition for cross-domain few-shot segmentation.arXiv preprint arXiv:2506.02677, 2025. 1

    2025 arXiv

  52. [60]

    Qwen3-vl technical report.arXiv preprint arXiv:2511.21631, 2025

    Shuai Bai, Yuxuan Cai, Ruizhe Chen, Keqin Chen, Xionghui Chen, Zesen Cheng, Lianghao Deng, Wei Ding, Chang Gao, Chunjiang Ge, et al. Qwen3-vl technical report.arXiv preprint arXiv:2511.21631, 2025. 3

    2025 arXiv

  53. [61]

    Llamafactory: Unified efficient fine-tuning of 100+ language models

    Yaowei Zheng, Richong Zhang, Junhao Zhang, Yanhan Ye, and Zheyan Luo. Llamafactory: Unified efficient fine-tuning of 100+ language models. InProceedings of the 62nd an- nual meeting of the association for computational linguistics (volume 3: system demonstrations), pages 400–4...

    2024

  54. [62]

    Omniego-r 2: A routed reasoning framework for the 1st cross-domain egocross chal- lenge at cvpr 2026.arXiv preprint arXiv:2605.24481, 2026

    Zixu Li, Zhiwei Chen, Zhiheng Fu, Wenbo Wang, Yupeng Hu, Weili Guan, and Liqiang Nie. Omniego-r 2: A routed reasoning framework for the 1st cross-domain egocross chal- lenge at cvpr 2026.arXiv preprint arXiv:2605.24481, 2026. 5

    2026 arXiv

  55. [63]

    The right inference strategy is all you need: Nearly training-free domain-wise inference for egocross challenge.arXiv preprint arXiv:2606.00829, 2026

    Leyi Wu, Yifan Zhao, Jinjie Zhang, Yinchuan Li, and Ying- Cong Chen. The right inference strategy is all you need: Nearly training-free domain-wise inference for egocross challenge.arXiv preprint arXiv:2606.00829, 2026. 4

    2026 arXiv

  56. [64]

    Reflective dialogue be- tween teacher and solver agents for video question answer- ing.arXiv preprint arXiv:2605.27885, 2026

    Takuya Murakawa and Toru Tamaki. Reflective dialogue be- tween teacher and solver agents for video question answer- ing.arXiv preprint arXiv:2605.27885, 2026. 6

    2026 arXiv

  57. [65]

    Gemini: a family of highly capable multimodal models.arXiv preprint arXiv:2312.11805, 2023

    Gemini Team, Rohan Anil, Sebastian Borgeaud, Jean- Baptiste Alayrac, Jiahui Yu, Radu Soricut, Johan Schalkwyk, Andrew M Dai, Anja Hauth, Katie Millican, et al. Gemini: a family of highly capable multimodal models.arXiv preprint arXiv:2312.11805, 2023. 7

    2023 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.