{"id":{"repo_id":"mit","oai_identifier":"oai:dspace.mit.edu:1721.1/145037"},"canonical_url":"https://search.dev.ndltd.org/etd/mit/oai:dspace.mit.edu:1721.1/145037","repository":{"repo_id":"mit","name":"MIT","base_url":"https://dspace.mit.edu/oai/request"},"display":{"title":"Measuring image difficulty under limited presentation time: towards building better test sets for object recognition","abstract":"Datasets are crucial to computer vision and broader machine learning. In particular, with the advance of techniques that are less well-understood theoretically, raw performance on datasets such as ImageNet has been the main driver of developments and feedback in the state of the field. However, the source of data that datasets draw on today are highly biased; for example, object class is correlated with backgrounds and omit many phenomena. In addition, objects mostly appear in stereotypical rotations with little occlusion. The resulting datasets themselves are similarly biased. Thus, the performance on datasets is limited as a predictor of the performance users can expect on their own tasks. To approach this problem, datasets such as ObjectNet were built with images that more closely resemble real-world scenarios by controlling for object backgrounds, rotations, imaging viewpoints, etc. In this thesis, we further address this problem by proposing a novel difficulty metric that reflects the performance of humans on recognizing images. We derive this metric by conducting extensive psychophysics experiments to determine the minimal time humans need to recognize an image. This new metric can be used to construct datasets that controls for the difficulty of different scenes and views humans see on a daily basis. The models’ performance on these datasets will also better represent the performance of humans on their own tasks. However, collecting these labels can be costly, so we also propose machine proxies that can effectively estimate human difficulty for different images and datasets.","abstract_html":"Datasets are crucial to computer vision and broader machine learning. In particular, with the advance of techniques that are less well-understood theoretically, raw performance on datasets such as ImageNet has been the main driver of developments and feedback in the state of the field. However, the source of data that datasets draw on today are highly biased; for example, object class is correlated with backgrounds and omit many phenomena. In addition, objects mostly appear in stereotypical rotations with little occlusion. The resulting datasets themselves are similarly biased. Thus, the performance on datasets is limited as a predictor of the performance users can expect on their own tasks. To approach this problem, datasets such as ObjectNet were built with images that more closely resemble real-world scenarios by controlling for object backgrounds, rotations, imaging viewpoints, etc. In this thesis, we further address this problem by proposing a novel difficulty metric that reflects the performance of humans on recognizing images. We derive this metric by conducting extensive psychophysics experiments to determine the minimal time humans need to recognize an image. This new metric can be used to construct datasets that controls for the difficulty of different scenes and views humans see on a daily basis. The models’ performance on these datasets will also better represent the performance of humans on their own tasks. However, collecting these labels can be costly, so we also propose machine proxies that can effectively estimate human difficulty for different images and datasets.","abstract_has_math":false,"creators":["Lin, Xinyu"],"institution":"Massachusetts Institute of Technology","degree_name":"Master","degree_level":null,"degree_discipline":null,"degree_department":"Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science","school":null,"contributors":[],"advisors":["Katz, Boris"],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-05","date_published":"2022-05","updated_at":"2026-07-22T22:20:53Z","subjects":[],"languages":[],"rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"rights_urls":["http://rightsstatements.org/page/InC-EDU/1.0/"],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/1721.1/145037","outbound_label":"Handle","outbound_source":"dc:identifier.uri"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor.advisor","label":"Advisor","values":["Katz, Boris"]},{"key":"dc:contributor.department","label":"Department","values":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"]},{"key":"dc:creator","label":"Author","values":["Lin, Xinyu"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date.accessioned","label":"Dc Date Accessioned","values":["2022-08-29T16:28:48Z"]},{"key":"dc:date.available","label":"Dc Date Available","values":["2022-08-29T16:28:48Z"]},{"key":"dc:date.issued","label":"Date","values":["2022-05"]},{"key":"dc:publisher","label":"Institution","values":["Massachusetts Institute of Technology"]},{"key":"dc:type","label":"Dc Type","values":["Thesis"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Master","Master of Engineering in Electrical Engineering and Computer Science"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:rights","label":"Dc Rights","values":["In Copyright - Educational Use Permitted","Copyright MIT"]},{"key":"dc:rights.uri","label":"Rights URI","values":["http://rightsstatements.org/page/InC-EDU/1.0/"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier.uri","label":"Identifier URI","values":["https://hdl.handle.net/1721.1/145037"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description.abstract","label":"Abstract","values":["Datasets are crucial to computer vision and broader machine learning. In particular, with the advance of techniques that are less well-understood theoretically, raw performance on datasets such as ImageNet has been the main driver of developments and feedback in the state of the field. However, the source of data that datasets draw on today are highly biased; for example, object class is correlated with backgrounds and omit many phenomena. In addition, objects mostly appear in stereotypical rotations with little occlusion. The resulting datasets themselves are similarly biased. Thus, the performance on datasets is limited as a predictor of the performance users can expect on their own tasks. To approach this problem, datasets such as ObjectNet were built with images that more closely resemble real-world scenarios by controlling for object backgrounds, rotations, imaging viewpoints, etc. In this thesis, we further address this problem by proposing a novel difficulty metric that reflects the performance of humans on recognizing images. We derive this metric by conducting extensive psychophysics experiments to determine the minimal time humans need to recognize an image. This new metric can be used to construct datasets that controls for the difficulty of different scenes and views humans see on a daily basis. The models’ performance on these datasets will also better represent the performance of humans on their own tasks. However, collecting these labels can be costly, so we also propose machine proxies that can effectively estimate human difficulty for different images and datasets."]},{"key":"dc:description.degree","label":"Dc Description Degree","values":["M.Eng."]},{"key":"dc:title","label":"Title","values":["Measuring image difficulty under limited presentation time: towards building better test sets for object recognition"]}]}],"canonical_facts":{"dc:contributor.advisor":["Katz, Boris"],"dc:contributor.department":["Massachusetts Institute of Technology. Department of Electrical Engineering and Computer Science"],"dc:creator":["Lin, Xinyu"],"dc:date.accessioned":["2022-08-29T16:28:48Z"],"dc:date.available":["2022-08-29T16:28:48Z"],"dc:date.issued":["2022-05"],"dc:description.abstract":["Datasets are crucial to computer vision and broader machine learning. In particular, with the advance of techniques that are less well-understood theoretically, raw performance on datasets such as ImageNet has been the main driver of developments and feedback in the state of the field. However, the source of data that datasets draw on today are highly biased; for example, object class is correlated with backgrounds and omit many phenomena. In addition, objects mostly appear in stereotypical rotations with little occlusion. The resulting datasets themselves are similarly biased. Thus, the performance on datasets is limited as a predictor of the performance users can expect on their own tasks. To approach this problem, datasets such as ObjectNet were built with images that more closely resemble real-world scenarios by controlling for object backgrounds, rotations, imaging viewpoints, etc. In this thesis, we further address this problem by proposing a novel difficulty metric that reflects the performance of humans on recognizing images. We derive this metric by conducting extensive psychophysics experiments to determine the minimal time humans need to recognize an image. This new metric can be used to construct datasets that controls for the difficulty of different scenes and views humans see on a daily basis. The models’ performance on these datasets will also better represent the performance of humans on their own tasks. However, collecting these labels can be costly, so we also propose machine proxies that can effectively estimate human difficulty for different images and datasets."],"dc:description.degree":["M.Eng."],"dc:identifier.uri":["https://hdl.handle.net/1721.1/145037"],"dc:publisher":["Massachusetts Institute of Technology"],"dc:rights":["In Copyright - Educational Use Permitted","Copyright MIT"],"dc:rights.uri":["http://rightsstatements.org/page/InC-EDU/1.0/"],"dc:title":["Measuring image difficulty under limited presentation time: towards building better test sets for object recognition"],"dc:type":["Thesis"],"thesis:degree_name":["Master","Master of Engineering in Electrical Engineering and Computer Science"]},"updated_at":"2026-07-22T22:20:53Z"}