{"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/what-uncertainties-do-we-need-in-bayesian","title":"What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision?","arxiv_id":"1703.04977","date":"2017-03-15","proceeding":"NeurIPS 2017 12","authors":["Alex Kendall","Yarin Gal"],"abstract":"There are two major types of uncertainty one can model. Aleatoric uncertainty\ncaptures noise inherent in the observations. On the other hand, epistemic\nuncertainty accounts for uncertainty in the model -- uncertainty which can be\nexplained away given enough data. Traditionally it has been difficult to model\nepistemic uncertainty in computer vision, but with new Bayesian deep learning\ntools this is now possible. We study the benefits of modeling epistemic vs.\naleatoric uncertainty in Bayesian deep learning models for vision tasks. For\nthis we present a Bayesian deep learning framework combining input-dependent\naleatoric uncertainty together with epistemic uncertainty. We study models\nunder the framework with per-pixel semantic segmentation and depth regression\ntasks. Further, our explicit uncertainty formulation leads to new loss\nfunctions for these tasks, which can be interpreted as learned attenuation.\nThis makes the loss more robust to noisy data, also giving new state-of-the-art\nresults on segmentation and depth regression benchmarks.","url_abs":"http://arxiv.org/abs/1703.04977v2","url_pdf":"http://arxiv.org/pdf/1703.04977v2.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":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/Glutamat42/epistemic_ensemble_toy_dataset","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/asharakeh/probdet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/hardianlawi/MTL-Homoscedastic-Uncertainty","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/hmi88/what","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/huyng/incertae","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok"}},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/kyle-dorman/bayesian-neural-network-blogpost","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/marcobellagente93/Bayesian_Regression","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/pmorerio/dl-uncertainty","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"MIT"}},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/taewankim1/uncertainty_deeplearning","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/tanyanair/segmentation_uncertainty","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":null},{"paper_slug":"what-uncertainties-do-we-need-in-bayesian","repo_url":"https://github.com/udion/DeepUncer","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok"}}],"tasks":[{"task_slug":"deep-learning","task_name":"Deep Learning"},{"task_slug":"depth-estimation","task_name":"Depth Estimation"},{"task_slug":"segmentation","task_name":"Segmentation"},{"task_slug":"semantic-segmentation","task_name":"Semantic Segmentation"},{"task_slug":"regression-1","task_name":"regression"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/semantic-segmentation-on-nyu-depth-v2","task":"Semantic Segmentation","dataset":"NYU Depth v2","model":"Bayesian DenseNet","rank_in_archive_order":114,"of":121,"metrics":{"Mean IoU":"37.3%"},"uses_additional_data":false}],"syntology":{"syntology_url":"https://syntology.ai/paper/1703.04977","atlas_url":"https://app.syntology.ai/?focus=1703.04977","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1703.04977"}},"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/asharakeh/probdet","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/pmorerio/dl-uncertainty","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/marcobellagente93/Bayesian_Regression","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/tanyanair/segmentation_uncertainty","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/kyle-dorman/bayesian-neural-network-blogpost","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hmi88/what","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/taewankim1/uncertainty_deeplearning","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/hardianlawi/MTL-Homoscedastic-Uncertainty","reach":{"status":"unanswered"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/Glutamat42/epistemic_ensemble_toy_dataset","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/udion/DeepUncer","reach":{"status":"ok"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/huyng/incertae","reach":{"status":"ok"}}],"summary":{"ran":4,"unverified":1},"by_repo_kind":{"listed":{"samples":5,"ran":4,"repositories":3}},"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":4,"samples":[{"code_sha256_prefix":"74f67ef7f3a7816f","entry":"COMBINED","repo":"hmi88/what","repo_kind":"listed","path":"WHAT_src/model/combined.py","file_url":"https://github.com/hmi88/what/blob/HEAD/WHAT_src/model/combined.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"74f67ef7f3a7816f"}},{"code_sha256_prefix":"52a4ba944c321852","entry":"LinearDensityNN","repo":"taewankim1/uncertainty_deeplearning","repo_kind":"listed","path":"models.py","file_url":"https://github.com/taewankim1/uncertainty_deeplearning/blob/HEAD/models.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"52a4ba944c321852"}},{"code_sha256_prefix":"2d58eb214c24d1e6","entry":"_Decoder","repo":"hmi88/what","repo_kind":"listed","path":"WHAT_src/model/combined.py","file_url":"https://github.com/hmi88/what/blob/HEAD/WHAT_src/model/combined.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2d58eb214c24d1e6"}},{"code_sha256_prefix":"a54ea00a730b6b46","entry":"_Encoder","repo":"hmi88/what","repo_kind":"listed","path":"WHAT_src/model/combined.py","file_url":"https://github.com/hmi88/what/blob/HEAD/WHAT_src/model/combined.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"a54ea00a730b6b46"}},{"code_sha256_prefix":"dfd995b09189de44","entry":"resize_images","repo":"pmorerio/dl-uncertainty","repo_kind":"listed","path":"prepro.py","file_url":"https://github.com/pmorerio/dl-uncertainty/blob/HEAD/prepro.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":"dfd995b09189de44"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}