Papers › Self-Chained Image-Language Model for Video Localization and Question Answering

Self-Chained Image-Language Model for Video Localization and Question Answering

11 May 2023NeurIPS 2023 11arXiv:2305.06988archive 2025-07-28

Shoubin Yu, Jaemin Cho, Prateek Yadav, Mohit Bansal

Recent studies have shown promising results on utilizing large pre-trained image-language models for video question answering. While these image-language models can efficiently bootstrap the representation learning of video-language models, they typically concatenate uniformly sampled video frames as visual inputs without explicit language-aware, temporal modeling. When only a portion of a video input is relevant to the language query, such uniform frame sampling can often lead to missing important visual cues. Although humans often find a video moment to focus on and rewind the moment to answer questions, training a query-aware video moment localizer often requires expensive annotations and high computational costs. To address this issue, we propose Self-Chained Video Localization-Answering (SeViLA), a novel framework that leverages a single image-language model (BLIP-2) to tackle both temporal keyframe localization and QA on videos. SeViLA framework consists of two modules: Localizer and Answerer, where both are parameter-efficiently fine-tuned from BLIP-2. We propose two ways of chaining these modules for cascaded inference and self-refinement. First, in the forward chain, the Localizer finds multiple language-aware keyframes in a video, which the Answerer uses to predict the answer. Second, in the reverse chain, the Answerer generates keyframe pseudo-labels to refine the Localizer, alleviating the need for expensive video moment localization annotations. Our SeViLA framework outperforms several strong baselines on 5 challenging video QA and event prediction benchmarks, and achieves the state-of-the-art in both fine-tuning (NExT-QA, STAR) and zero-shot (NExT-QA, STAR, How2QA, VLEP) settings. We also analyze the impact of Localizer, comparisons of Localizer with other temporal localization models, pre-training/self-refinement of Localizer, and varying the number of keyframes.

PaperPDFConference PDFCodeCode Syntology ran

In Syntology Open this paper in Syntology's Atlas, the map of the papers in Syntology's graph and their citations.

For agents, Syntology's MCP tool lists every function and class Syntology harvested from this paper and whether it ran (how to connect): get_harvested_code_for_paper(arxiv_id="2305.06988")

Code

Syntology Ran 6 of 7 code samples harvested from 1 repository linked to this paper; 1 has no recorded run. Of those that ran: 6 ran with no contract checked.

By repository: official repository: 7 samples from 1 repository, 6 ran. The run record, sample by sample. “Ran” means executed on a synthesized input, not that the code is correct or reproduces the paper.

yui010206/sevila officialmentioned in papermentioned on GitHubpytorchBSD-3-Clause report

Repository list and official/mentioned flags are the archive's, frozen 2025-07-28. Reachability, where shown, is from one Syntology probe window (2026-09-16 to 2026-09-18); repositories not probed show nothing. GitHub stars are not tracked.

Code Syntology ran Syntology

7 samples harvested; 6 ran; 0 honoured the contract we drafted; 1 has no recorded run. Read from Syntology's graph 2026-09-24; that is when this build read the record, not when the samples ran.

6ran
1unverified

Licence: 7 of the 7 samples are pointer only, meaning Syntology does not serve that copy's text. This page shows no code text for any sample; each one links to its file in the repository.

Harvested from yui010206/sevila. “Ran” means the sample executed on a synthesized input. It does not mean the output is correct, and nothing here reproduces the paper's results. “Honoured” and “violated” refer to a contract Syntology drafted from the code itself; “our draft was wrong” and “fixture could not drive it” are failures of Syntology's instrument, not of the code.

Each sample ends with its code_sha256, Syntology's identity for that exact code. An agent fetches the stored sample with Syntology's MCP tool get_code(code_sha256="…") (how to connect); click an identity to copy that call.

Repository labels, per sample. official repository: The archive marks this repository official for the paper. named in the paper: The archive records that the paper mentions this repository; it is not marked official. community (archive-listed): In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper. found in paper text by Syntology: Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted. community: Not in the archive's code links for this paper; a community repository Syntology harvested. Samples from a repository marked official are listed first. Licence labels name the repository's licence as recorded at harvest. “Pointer only” means Syntology does not serve that copy's text, for one of four reasons: no licence file was found; the licence was not identified; the licence is recorded as permissive but that copy's record is not marked cleared; or the licence is outside the permissive list Syntology serves text under (MIT, Apache-2.0, BSD and similar). Some licences outside that list permit redistribution, such as WTFPL, and GPL-3.0 under its conditions; they are simply not on the list. Hover a licence label for the reason. File links open the file on GitHub at the default branch, which may have changed since the harvest.

generate_caption yui010206/sevila/app/caption.py official repository ran BSD-3-Clause recorded; this copy not marked cleared · pointer only · 56f02812d66a0d11 · report
get_concat_v yui010206/sevila/app/dataset_browser.py official repository ran BSD-3-Clause recorded; this copy not marked cleared · pointer only · 92f2e16ee0d3a24d · report
read_img yui010206/sevila/app/calculate_coco_features.py official repository ran BSD-3-Clause recorded; this copy not marked cleared · pointer only · 13a43a7d815937e6 · report
resize_img yui010206/sevila/app/utils.py official repository ran BSD-3-Clause recorded; this copy not marked cleared · pointer only · 926bb38f979a61ef · report
resize_img_w yui010206/sevila/app/dataset_browser.py official repository ran BSD-3-Clause recorded; this copy not marked cleared · pointer only · f809f2ecf34c1c99 · report
sample_dataset yui010206/sevila/app/dataset_browser.py official repository ran BSD-3-Clause recorded; this copy not marked cleared · pointer only · 3d4bbb5ca53fd24a · report
compute_gradcam_batch yui010206/sevila/app/multimodal_search.py official repository unverified BSD-3-Clause recorded; this copy not marked cleared · pointer only · 3e12ea8c09674aa8 · report

Tasks

Language ModelingLanguage ModellingQuestion AnsweringRepresentation LearningTemporal LocalizationVideo Question AnsweringZero-Shot Video Question Answer

Results from the paper archive 2025-07-28

TaskDatasetModelMetricValueRank at snapshotLeaderboardReport
Video Question Answering NExT-QA SeViLA Accuracy 73.8 #27 of 47 Archive leaderboard report
Video Question Answering NExT-QA (Efficient) SeViLA (4 frames) 1:1 Accuracy 73.8 #2 of 2 Archive leaderboard report
Video Question Answering STAR Benchmark SeViLA Average Accuracy 64.9 #3 of 17 Archive leaderboard report
Video Question Answering STAR Benchmark SeViLA (0-shot) Average Accuracy 44.6 #12 of 17 Archive leaderboard report
Zero-Shot Video Question Answer EgoSchema (fullset) SeViLA (4B) Accuracy 22.7 #28 of 29 Archive leaderboard report
Zero-Shot Video Question Answer EgoSchema (subset) SeViLA (4B) Accuracy 25.7 #13 of 14 Archive leaderboard report
Zero-Shot Video Question Answer IntentQA SeViLA (4B) Accuracy 60.9 #8 of 13 Archive leaderboard report
Zero-Shot Video Question Answer NExT-QA Sevila (4B) Accuracy 63.6 #19 of 27 Archive leaderboard report
Zero-Shot Video Question Answer TVQA SEVILA (no speech) Accuracy 38.2 #7 of 9 Archive leaderboard report

Ranks are positions in the archive's leaderboards as they stood at the 2025-07-28 snapshot. Results published since then are not among these rows, so a rank here is not a current standing.

Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections