Pith. sign in

REVIEW 3 major objections 5 minor 30 references

VidCtx: Context-aware Video Question Answering with Image Models

T0 review · 3 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read VidCtx claims a training-free recipe that pairs each video frame with a question-aware caption of a distant frame and max-pools the per-frame answers, reaching the best open-model results on NExT-QA and IntentQA.

desk verdict VidCtx is a sound, training-free VideoQA pipeline whose internal ablations are convincing; the reported SOTA margin is plausible but not fully controlled because the baselines were not re-run. read the letter →

arxiv 2412.17415 v2 pith:D4HIWLQQ submitted 2024-12-23 cs.CV cs.AIcs.MM

Andreas Goulas , Vasileios Mezaris , Ioannis Patras This is my paper · ORCID
classification cs.CVcs.AIcs.MM
keywords largemultimodalmodelsvideoquestionansweringzero-shotlearningquestion-awarecaptionsframe-levelvotingmaxpoolingtemporalreasoningmultiple-choiceQA
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 paper argues that a single off-the-shelf image-language model can answer multiple-choice video questions well without any video training, if each sampled frame is paired with a question-aware caption of a distant frame and the per-frame choices are combined by max pooling. The claim is that visual and textual modalities are complementary, and that a caption placed half a video away gives the model enough temporal context to reason about causal and temporal questions. On NExT-QA and IntentQA this recipe reaches 70.7% and 67.1% accuracy, the best among open-model methods in the paper's comparison, and it stays competitive on STAR. The practical interest is that the approach is training-free, uses one 7B LMM for both captioning and answering, and scales linearly with the number of frames rather than quadratically.

What carries the argument

The load-bearing mechanism is the distant-frame pairing: for each sampled frame $i$ of $N$, the answer prompt receives the question-aware caption of frame $r(i) \equiv i + \frac{N}{2} \pmod{N}$, together with an 'earlier' or 'later' temporal specifier. Captions are produced by prompting the same LMM with a short description instruction that includes the question. A special 'No Answer' token lets the model abstain on irrelevant frames, and the final decision is $y = \arg\max_{t \in T-\{F\}} \max_{0 \le i < N} \frac{p(d_i=t)}{\sum_{k\in T} p(d_i=k)}$, i.e. L1-normalized per-frame first-token probabilities passed through max pooling. This mechanism converts the temporal structure of the video into local pairwise constraints, and it is what lets the method scale to an arbitrary number of frames.

What would settle it

Re-run Q-ViD's caption-concatenation pipeline with the same LLaVa-1.6-Mistral-7B backbone, 4-bit quantization, 64-frame sampling, the same question-aware captions, and the same answer prompt; if its NExT-QA accuracy reaches or exceeds 70.7%, the paper's claimed +4.4% advantage over Q-ViD would disappear.

Watch

Extended reading notes

Core claim

VidCtx's central claim is that zero-shot multiple-choice video question answering improves when the answer about a frame is conditioned on the question-aware caption of a second frame located roughly half a video away, and when the resulting per-frame answer distributions are aggregated by L1-normalized max pooling. The paper reports 70.7% top-1 on NExT-QA, which is +4.4% over Q-ViD and the best among the open-model methods in its table; 67.1% on IntentQA, +3.5% over Q-ViD; and 51.1% average on STAR, second to VideoChat2, a model pre-trained on large video corpora. VidCtx also reports beating the GPT-3.5-based LLoVi pipeline while using a 7B open model. The authors take this as evidence that combining the visual signal of one frame with text from a distant frame captures temporal structure better than either captions-only or single-frame visual processing alone.

Load-bearing premise

The reported gains over Q-ViD, SeViLA, LLoVi, and LangRepo come from numbers published in those papers and not from re-running them under VidCtx's exact settings; if those models were evaluated with the same frame count, quantization, caption budget, and LLaVa backbone, the margins could change.

Editorial extensions

If this is right

  • Scaling frames raises accuracy monotonically on NExT-QA, from 63.5% with one frame to 70.7% with 64 frames, and the frame-by-frame design removes the context-length ceiling that limits caption-concatenation methods.
  • Using a question-aware distant caption adds +2.8 points over no context and +1.2 points over static distant captions, so context relevance to the question matters, not just context presence.
  • The same visual-plus-caption setup beats a captions-only baseline by +3.0 points when both use identical captions and the same model, supporting the paper's claim that visual and textual modalities are complementary.
  • Because nothing is trained, the pipeline can run on any single-image instruction-tuned LMM and any choice of frame count, making it a drop-in recipe for zero-shot video QA.

Reading between the lines

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

  • An untested but direct extension is to choose the paired frame adaptively, for example by caption similarity to the question or by the model's abstention signal, instead of always half a video away; the paper's own context ablations suggest the optimum may vary with video length.
  • The +3.5 to +4.4 point margins over Q-ViD rely on published baseline numbers; re-running those baselines under VidCtx's exact sampling, 4-bit quantization, and LLaVa backbone would show how much of the gap is method versus configuration.
  • The architecture implies that any improved single-image LMM could be converted into a video QA system with no additional training, so the practical value of the recipe may transfer to stronger image models as they are released.
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

3 major / 5 minor

Summary. The paper proposes VidCtx, a training-free video question-answering framework that combines visual evidence from individual frames with question-aware textual captions of distant frames. A pre-trained image LMM (LLaVA-1.6-Mistral-7B, 4-bit) is used both to generate captions for sampled frames and to produce frame-level multiple-choice decisions conditioned on the frame, the question, and a distant-frame caption. The frame-level decisions are aggregated by max pooling of L1-normalized first-token scores. The method is evaluated zero-shot on NExT-QA, IntentQA, and STAR, where it reports strong performance among open-model approaches. The paper contributes a simple, modular architecture, careful ablations of context choice and aggregation, and a qualitative analysis.

Significance. If the reported results hold, VidCtx is a valuable contribution to training-free VideoQA: it demonstrates that combining per-frame visual input with textual context from distant frames is complementary, and it does so with a cheap, reproducible pipeline that scales linearly in the number of frames. The internal ablations (Tables II, III, IV, V) are well designed and support the key qualitative claims (distant question-aware captions help; max pooling with L1 normalization helps). The release of code is a further strength. The main reservation is that the headline state-of-the-art comparison against Q-ViD, SeViLA, and other open approaches in Table I relies on numbers taken from their original papers under different backbones, frame counts, and prompts, so the exact margins are not fully controlled. The central mechanism, however, is supported by the same-model ablations.

major comments (3)
  1. [Section IV-C, Table I] The claim that VidCtx achieves state-of-the-art performance among open approaches on NExT-QA and IntentQA is based on accuracy numbers copied from the original papers of Q-ViD, SeViLA, LLoVi, and others. Those methods use different backbone LMMs (e.g., Q-ViD is 12B with a separate captioner and reasoner), different frame sampling strategies, and different prompt templates. Since VidCtx uses 4-bit LLaVA-1.6-Mistral-7B with 64 frames on NExT-QA/IntentQA and 32 on STAR, the reported +4.4% and +3.5% margins over Q-ViD may be partly attributable to the stronger base model rather than to the proposed context mechanism. To support the SOTA claim, the authors should re-run at least the most competitive open baselines under VidCtx's exact settings (same LMM, same frame count, same prompt), or explicitly qualify the comparison as non-controlled and use wording such as 'competitive' rather than 'state-of-the-art'.
  2. [Section III-D, Eq. (4)] Equation (4) is the core aggregation mechanism, but its notation is ambiguous and potentially incorrect. The text in Section III-D says 'we consider the probability of just the first token,' yet Eq. (4) defines p(d_i = t) as a 'log probability score.' If p is a log probability, it is typically negative, and the L1 normalization in the denominator sums absolute values, producing normalized scores that can be negative. This is an unusual and unjustified operation. If instead p denotes a probability, the denominator should sum probabilities, not their absolute values. The authors should clarify exactly which quantity is used, how the first token is extracted (e.g., after prompting the model to output a single letter), and why the absolute-value L1 normalization is appropriate. This is essential for reproducibility of the central max-pooling mechanism.
  3. [Section IV-D, Tables II-V] All ablation results are reported as single-run top-1 accuracies without variance or significance testing. Several differences that drive design choices are small (e.g., 70.7 vs. 70.3 for distant vs. current caption in Table II; 70.7 vs. 70.6 for L1+max vs. softmax+max in Table V). If decoding is stochastic (e.g., temperature sampling), these differences could be within run-to-run noise. The authors should either report multiple runs with different seeds, provide evidence that inference is deterministic, or explicitly state that these differences are qualitative rather than statistically established.
minor comments (5)
  1. [Section III-C, Eq. (2)] Equation (2) writes d_i = φ(v_i, c_{r(i)}, I_vqa), but I_vqa includes the question Q; this dependency is clear in the text but would be clearer if the equation also showed Q explicitly, for instance d_i = φ(v_i, c_{r(i)}, Q, I_vqa).
  2. [Section III-D] The special token F ('No Answer') is excluded from the argmax in Eq. (4), but the paper does not specify what happens if the LMM outputs F for every frame. A fallback policy (e.g., choose the highest-scoring option despite F, or predict a default) should be stated.
  3. [Section IV-A] The sentence 'We generate 200 tokens per caption' should specify whether 200 is the maximum generation length or an exact target length; otherwise the captioning cost is not precisely described.
  4. [Section IV-C, Computational Complexity] The statement that related works such as LLoVi and Q-ViD 'scale quadratically' due to attention in the LLM is a simplification; the quadratic scaling is in the number of tokens, and those methods may truncate or chunk captions. A more precise description of the complexity comparison would be helpful.
  5. [Section IV-D, Table V] The aggregation method labeled 'Voting' is not defined in the text. Does it mean majority vote over the argmax of each frame's answer, or something else? Please define it.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: VidCtx is an empirical pipeline whose components are independently justified by same-model ablations; the external comparison issue is a benchmarking limitation, not circular reasoning.

full rationale

The paper is an empirical systems paper. VidCtx's pipeline extracts question-aware captions, generates frame-level decisions with the same LMM using distant captions as context, and max-pools normalized per-frame answer probabilities. Each component is an explicit construction (Eqs. 1-4), and the final decision is a function of frame pixels, the question, and captions; it is not defined in terms of the target answer, nor is any fitted parameter renamed as a prediction. The internal ablations (Tables II, III, IV, V) compare variants under the same model and captions, so the claimed benefit of distant question-aware context is not an artifact of construction. The reported SOTA comparisons copy baseline numbers from prior papers under different settings (frame counts, prompts, base LMMs); however, this is a comparison-fairness or benchmark-reproduction limitation, not circular reasoning, because the paper's central mechanism is independently supported by its same-model ablations. The method follows Q-ViD's captioning and prompt choices via citation, but that is an external, non-self citation and does not import an unverified uniqueness claim or define the result in terms of itself. No load-bearing self-citation, uniqueness-theorem chain, or self-definitional step exists in the manuscript.

Assumptions & free parameters 4 free parameters · 3 assumptions · 0 invented entities

The method introduces no new theoretical objects. Its load-bearing choices are empirical hyperparameters (frame count, caption length, distant offset, normalization) and two behavioral assumptions about how the LMM responds to captions and first-token probabilities. All are selected via validation-set ablations, so the reported numbers partly reflect fitting to those validation sets.

free parameters (4)
  • Number of sampled frames N = 64 for NExT-QA/IntentQA, 32 for STAR
    Chosen per dataset based on average video duration; ablations in Table IV show accuracy increases with N, so the reported numbers use the largest feasible N.
  • Caption length limit = 200 tokens
    Chosen to give detailed question-aware captions; no ablation provided.
  • Distant frame offset = N/2 modulo N
    Selected to maximize temporal diversity between frame and context; no ablation comparing other offsets.
  • Normalization and pooling = L1 norm + max pooling
    Selected by validation-set comparison in Table V; differences among pooling methods are small (0.4-1.4 points).
assumptions (3)
  • domain assumption Uniform sampling of one central frame per segment yields a representative view of each video segment.
    Stated in Section III-B; unverified across datasets.
  • domain assumption The LMM's first-token log probabilities across answer options are a reliable basis for frame-level decisions.
    Implicit in Section III-D; the model's uncertainty is read directly from a single token.
  • domain assumption Appending a caption of a distant frame with an earlier/later specifier lets the LMM reason about temporal order.
    Core design choice in Section III-C; its effectiveness is shown only indirectly through ablations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of VidCtx: Context-aware Video Question Answering with Image Models." pith.science (2026). https://pith.science/paper/D4HIWLQQ

@misc{pith2026241217415,
  author       = {Pith},
  title        = {Pith review of: VidCtx: Context-aware Video Question Answering with Image Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/D4HIWLQQ}},
  note         = {Machine review of arXiv:2412.17415}
}
read the original abstract

To address computational and memory limitations of Large Multimodal Models in the Video Question-Answering task, several recent methods extract textual representations per frame (e.g., by captioning) and feed them to a Large Language Model (LLM) that processes them to produce the final response. However, in this way, the LLM does not have access to visual information and often has to process repetitive textual descriptions of nearby frames. To address those shortcomings, in this paper, we introduce VidCtx, a novel training-free VideoQA framework which integrates both modalities, i.e. both visual information from input frames and textual descriptions of others frames that give the appropriate context. More specifically, in the proposed framework a pre-trained Large Multimodal Model (LMM) is prompted to extract at regular intervals, question-aware textual descriptions (captions) of video frames. Those will be used as context when the same LMM will be prompted to answer the question at hand given as input a) a certain frame, b) the question and c) the context/caption of an appropriate frame. To avoid redundant information, we chose as context the descriptions of distant frames. Finally, a simple yet effective max pooling mechanism is used to aggregate the frame-level decisions. This methodology enables the model to focus on the relevant segments of the video and scale to a high number of frames. Experiments show that VidCtx achieves competitive performance among approaches that rely on open models on three public Video QA benchmarks, NExT-QA, IntentQA and STAR. Our code is available at https://github.com/IDT-ITI/VidCtx.

Figures

Figures reproduced from arXiv: 2412.17415 by the authors.

Figure 1
Figure 1. Comparison of video understanding frameworks. Related works typically either project the input frames to the LLM input space or rely on extracted captions to answer questions about videos. We propose combining both modalities, processing the video frame-by-frame and inserting the relevant context (i.e., extracted question-aware captions) as part of the LLM prompt. In our architecture, we use the same LMM for both ca… view at source ↗
Figure 2
Figure 2. Qualitative study of VidCtx on NExT-QA. We select one question from each question category and we present the normalized answer scores for pairs of distant frames, along with their question aware captions. [11] Ce Zhang, Taixi Lu, Md Mohaiminul Islam, Ziyang Wang, Shoubin Yu, et al., “A simple llm framework for long-range video question￾answering,” arXiv preprint arXiv:2312.17235, 2023. [12] Kumara Kahatapitiya, Kan… view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

30 extracted references · 16 canonical work pages

  1. [1]

    Video-chatgpt: Towards detailed video understanding via large vision and language models,

    Muhammad Maaz, Hanoona Rasheed, Salman Khan, and Fahad Shahbaz Khan, “Video-chatgpt: Towards detailed video understanding via large vision and language models,” arXiv preprint arXiv:2306.05424 , 2023

    arXiv 2023

  2. [2]

    Video-llama: An instruction- tuned audio-visual language model for video understanding,

    Hang Zhang, Xin Li, and Lidong Bing, “Video-llama: An instruction- tuned audio-visual language model for video understanding,” arXiv preprint arXiv:2306.02858, 2023

    arXiv 2023

  3. [3]

    Video-llava: Learning united visual representation by alignment before projection,

    Bin Lin, Bin Zhu, Yang Ye, Munan Ning, Peng Jin, and Li Yuan, “Video-llava: Learning united visual representation by alignment before projection,” arXiv preprint arXiv:2311.10122 , 2023

    arXiv 2023

  4. [4]

    Video instruction tuning with synthetic data,

    Yuanhan Zhang, Jinming Wu, Wei Li, Bo Li, Zejun Ma, Ziwei Liu, and Chunyuan Li, “Video instruction tuning with synthetic data,” arXiv preprint arXiv:2410.02713, 2024

    arXiv 2024

  5. [5]

    Mvbench: A comprehensive multi-modal video understanding bench- mark,

    Kunchang Li, Yali Wang, Yinan He, Yizhuo Li, Yi Wang, et al., “Mvbench: A comprehensive multi-modal video understanding bench- mark,” in Proc. IEEE/CVF CVPR , 2024, pp. 22195–22206

    work page 2024

  6. [6]

    Qwen2-vl: Enhancing vision-language model’s perception of the world at any resolution,

    Peng Wang, Shuai Bai, Sinan Tan, Shijie Wang, Zhihao Fan, et al., “Qwen2-vl: Enhancing vision-language model’s perception of the world at any resolution,” arXiv preprint arXiv:2409.12191 , 2024

    arXiv 2024

  7. [7]

    Internvideo2: Scaling video foundation models for multimodal video understanding,

    Yi Wang, Kunchang Li, Xinhao Li, Jiashuo Yu, Yinan He, et al., “Internvideo2: Scaling video foundation models for multimodal video understanding,” arXiv preprint arXiv:2403.15377 , 2024

    arXiv 2024

  8. [8]

    Self- chained image-language model for video localization and question answering,

    Shoubin Yu, Jaemin Cho, Prateek Yadav, and Mohit Bansal, “Self- chained image-language model for video localization and question answering,” Advances in Neural Information Processing Systems , vol. 36, 2023

    work page 2023

Show all 30 references
  1. [9]

    Videotree: Adaptive tree-based video representation for llm reasoning on long videos,

    Ziyang Wang, Shoubin Yu, Elias Stengel-Eskin, Jaehong Yoon, Feng Cheng, et al., “Videotree: Adaptive tree-based video representation for llm reasoning on long videos,” arXiv preprint arXiv:2405.19209, 2024

    2024 arXiv

  2. [10]

    Videoagent: Long-form video understanding with large language model as agent,

    Xiaohan Wang, Yuhui Zhang, Orr Zohar, and Serena Yeung-Levy, “Videoagent: Long-form video understanding with large language model as agent,” in Proc. ECCV. Springer, 2025, pp. 58–76. Question: What did the white dog do after he looked up? (Category: Temporal) Frame 20 of 64 Fr...

    2025

  3. [11]

    A simple llm framework for long-range video question- answering,

    Ce Zhang, Taixi Lu, Md Mohaiminul Islam, Ziyang Wang, Shoubin Yu, et al., “A simple llm framework for long-range video question- answering,” arXiv preprint arXiv:2312.17235 , 2023

    2023 arXiv

  4. [12]

    Language repository for long video understanding,

    Kumara Kahatapitiya, Kanchana Ranasinghe, Jongwoo Park, and Michael S Ryoo, “Language repository for long video understanding,” arXiv preprint arXiv:2403.14622 , 2024

    2024 arXiv

  5. [13]

    Question-instructed visual de- scriptions for zero-shot video answering,

    David Mogrovejo and Thamar Solorio, “Question-instructed visual de- scriptions for zero-shot video answering,” in Findings of the Association for Computational Linguistics ACL 2024 , 2024, pp. 9329–9339

    2024

  6. [14]

    Same task, more tokens: the impact of input length on the reasoning performance of large language models,

    Mosh Levy, Alon Jacoby, and Yoav Goldberg, “Same task, more tokens: the impact of input length on the reasoning performance of large language models,” arXiv preprint arXiv:2402.14848 , 2024

    2024 arXiv

  7. [15]

    Large language models can be easily distracted by irrelevant context,

    Freda Shi, Xinyun Chen, Kanishka Misra, Nathan Scales, David Dohan, et al., “Large language models can be easily distracted by irrelevant context,” in Int. Conf. on Machine Learning . PMLR, 2023, pp. 31210– 31227

    2023

  8. [16]

    Learn to explain: Multimodal reasoning via thought chains for science question answering,

    Pan Lu, Swaroop Mishra, Tanglin Xia, Liang Qiu, Kai-Wei Chang, et al., “Learn to explain: Multimodal reasoning via thought chains for science question answering,” Advances in Neural Information Processing Systems, vol. 35, pp. 2507–2521, 2022

    2022

  9. [17]

    Multimodal chain-of-thought reasoning in language models,

    Zhuosheng Zhang, Aston Zhang, Mu Li, Hai Zhao, George Karypis, and Alex Smola, “Multimodal chain-of-thought reasoning in language models,” arXiv preprint arXiv:2302.00923 , 2023

    2023 arXiv

  10. [18]

    Ddcot: Duty-distinct chain-of-thought prompting for multimodal rea- soning in language models,

    Ge Zheng, Bin Yang, Jiajin Tang, Hong-Yu Zhou, and Sibei Yang, “Ddcot: Duty-distinct chain-of-thought prompting for multimodal rea- soning in language models,” Advances in Neural Information Processing Systems, vol. 36, pp. 5168–5191, 2023

    2023

  11. [19]

    Enhancing multimodal sentiment analysis via learning from large language model,

    Ning Pang, Wansen Wu, Yue Hu, Kai Xu, Quanjun Yin, and Long Qin, “Enhancing multimodal sentiment analysis via learning from large language model,” in Proc. IEEE ICME , 2024, pp. 1–6

    2024

  12. [20]

    Video-of-thought: Step-by-step video reasoning from perception to cognition,

    Hao Fei, Shengqiong Wu, Wei Ji, Hanwang Zhang, Meishan Zhang, Mong-Li Lee, and Wynne Hsu, “Video-of-thought: Step-by-step video reasoning from perception to cognition,” in Forty-first International Conference on Machine Learning , 2024

    2024

  13. [21]

    Vamos: Versatile action models for video understanding,

    Shijie Wang, Qi Zhao, Minh Quan Do, Nakul Agarwal, Kwonjoon Lee, and Chen Sun, “Vamos: Versatile action models for video understanding,” arXiv preprint arXiv:2311.13627 , 2023

    2023 arXiv

  14. [22]

    Large language models are zero-shot reasoners,

    Takeshi Kojima, Shixiang Shane Gu, Machel Reid, Yutaka Matsuo, and Yusuke Iwasawa, “Large language models are zero-shot reasoners,” Advances in neural information processing systems , vol. 35, pp. 22199– 22213, 2022

    2022

  15. [23]

    Chain-of-thought prompting elicits reasoning in large language models,

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, et al., “Chain-of-thought prompting elicits reasoning in large language models,” Advances in neural information processing systems , vol. 35, pp. 24824–24837, 2022

    2022

  16. [24]

    Compositional chain-of-thought prompting for large multimodal mod- els,

    Chancharik Mitra, Brandon Huang, Trevor Darrell, and Roei Herzig, “Compositional chain-of-thought prompting for large multimodal mod- els,” in Proc. IEEE/CVF CVPR , 2024, pp. 14420–14431

    2024

  17. [25]

    Next-qa: Next phase of question-answering to explaining temporal actions,

    Junbin Xiao, Xindi Shang, Angela Yao, and Tat-Seng Chua, “Next-qa: Next phase of question-answering to explaining temporal actions,” in Proc. IEEE/CVF CVPR , 2021, pp. 9777–9786

    2021

  18. [26]

    Intentqa: Context- aware video intent reasoning,

    Jiapeng Li, Ping Wei, Wenjuan Han, and Lifeng Fan, “Intentqa: Context- aware video intent reasoning,” in Proc. IEEE/CVF ICCV , 2023, pp. 11963–11974

    2023

  19. [27]

    Star: A benchmark for situated reasoning in real-world videos,

    Bo Wu, Shoubin Yu, Zhenfang Chen, Joshua B Tenenbaum, and Chuang Gan, “Star: A benchmark for situated reasoning in real-world videos,” arXiv preprint arXiv:2405.09711 , 2024

    2024 arXiv

  20. [28]

    Verbs in action: Improving verb understanding in video-language models,

    Liliane Momeni, Mathilde Caron, Arsha Nagrani, Andrew Zisserman, and Cordelia Schmid, “Verbs in action: Improving verb understanding in video-language models,” in Proc. IEEE/CVF ICCV, 2023, pp. 15579– 15591

    2023

  21. [29]

    Internvideo: General video foundation models via generative and dis- criminative learning,

    Yi Wang, Kunchang Li, Yizhuo Li, Yinan He, Bingkun Huang, et al., “Internvideo: General video foundation models via generative and dis- criminative learning,” arXiv preprint arXiv:2212.03191 , 2022

    2022 arXiv

  22. [30]

    Llava-next: Improved reasoning, ocr, and world knowledge,

    Haotian Liu, Chunyuan Li, Yuheng Li, Bo Li, Yuanhan Zhang, Sheng Shen, and Yong Jae Lee, “Llava-next: Improved reasoning, ocr, and world knowledge,” January 2024, Accessed: Dec. 12, 2024. [Online]. Available: https://llava-vl.github.io/blog/2024-01-30-llava-next/

    2024

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.