{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/a-dataset-and-exploration-of-models-for","title":"A dataset and exploration of models for understanding video data through fill-in-the-blank question-answering","arxiv_id":"1611.07810","date":"2016-11-23","proceeding":"CVPR 2017 7","authors":["Tegan Maharaj","Nicolas Ballas","Anna Rohrbach","Aaron Courville","Christopher Pal"],"abstract":"While deep convolutional neural networks frequently approach or exceed\nhuman-level performance at benchmark tasks involving static images, extending\nthis success to moving images is not straightforward. Having models which can\nlearn to understand video is of interest for many applications, including\ncontent recommendation, prediction, summarization, event/object detection and\nunderstanding human visual perception, but many domains lack sufficient data to\nexplore and perfect video models. In order to address the need for a simple,\nquantitative benchmark for developing and understanding video, we present\nMovieFIB, a fill-in-the-blank question-answering dataset with over 300,000\nexamples, based on descriptive video annotations for the visually impaired. In\naddition to presenting statistics and a description of the dataset, we perform\na detailed analysis of 5 different models' predictions, and compare these with\nhuman performance. We investigate the relative importance of language, static\n(2D) visual features, and moving (3D) visual features; the effects of\nincreasing dataset size, the number of frames sampled; and of vocabulary size.\nWe illustrate that: this task is not solvable by a language model alone; our\nmodel combining 2D and 3D visual information indeed provides the best result;\nall models perform significantly worse than human-level. We provide human\nevaluations for responses given by different models and find that accuracy on\nthe MovieFIB evaluation corresponds well with human judgement. We suggest\navenues for improving video models, and hope that the proposed dataset can be\nuseful for measuring and encouraging progress in this very interesting field.","url_abs":"http://arxiv.org/abs/1611.07810v2","url_pdf":"http://arxiv.org/pdf/1611.07810v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"a-dataset-and-exploration-of-models-for","repo_url":"https://github.com/teganmaharaj/movieFIB","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"a-dataset-and-exploration-of-models-for","repo_url":"https://github.com/totalgood/viddesc","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"descriptive","task_name":"Descriptive"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"object-detection","task_name":"Object Detection"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"object-detection-1","task_name":"object-detection"}],"methods":[],"datasets_introduced":[{"slug":"moviefib","name":"MovieFIB","full_name":"Movie Fill-in-the-Blank"}],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1611.07810","atlas_url":"https://app.syntology.ai/?focus=1611.07810","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1611.07810"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/totalgood/viddesc","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/teganmaharaj/movieFIB","reach":{"status":"ok"}}],"summary":{"unverified":8},"by_repo_kind":{"listed":{"samples":8,"ran":0,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"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"},"n_pointer_only_for_licence":0,"samples":[{"code_sha256_prefix":"9e6c2c92eee0a29e","entry":"bytes_feature","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/viddesc/preprocess.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/viddesc/preprocess.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9e6c2c92eee0a29e"}},{"code_sha256_prefix":"c39b7423940b40e6","entry":"collate_fn","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/imgdesc/data_loader.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/imgdesc/data_loader.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c39b7423940b40e6"}},{"code_sha256_prefix":"d3d0bc826c2e6782","entry":"define_models","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/viddesc/keras_lstm.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/viddesc/keras_lstm.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"d3d0bc826c2e6782"}},{"code_sha256_prefix":"999ee8d595b9fdf7","entry":"fib","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/viddesc/skeleton.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/viddesc/skeleton.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"999ee8d595b9fdf7"}},{"code_sha256_prefix":"44fda9771f751cfc","entry":"float_feature","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/viddesc/preprocess.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/viddesc/preprocess.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"44fda9771f751cfc"}},{"code_sha256_prefix":"0c2d6e281d5fbeb5","entry":"int64_feature","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/viddesc/preprocess.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/viddesc/preprocess.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0c2d6e281d5fbeb5"}},{"code_sha256_prefix":"5933f2701ea63955","entry":"load_embeddings","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/viddesc/keras_lstm.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/viddesc/keras_lstm.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5933f2701ea63955"}},{"code_sha256_prefix":"ddccddb3e642d270","entry":"resize_image","repo":"totalgood/viddesc","repo_kind":"listed","path":"src/imgdesc/resize.py","file_url":"https://github.com/totalgood/viddesc/blob/HEAD/src/imgdesc/resize.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ddccddb3e642d270"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}