{"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/190503711","title":"Processing Megapixel Images with Deep Attention-Sampling Models","arxiv_id":"1905.03711","date":"2019-05-03","proceeding":null,"authors":["Angelos Katharopoulos","François Fleuret"],"abstract":"Existing deep architectures cannot operate on very large signals such as megapixel images due to computational and memory constraints. To tackle this limitation, we propose a fully differentiable end-to-end trainable model that samples and processes only a fraction of the full resolution input image. The locations to process are sampled from an attention distribution computed from a low resolution view of the input. We refer to our method as attention sampling and it can process images of several megapixels with a standard single GPU setup. We show that sampling from the attention distribution results in an unbiased estimator of the full model with minimal variance, and we derive an unbiased estimator of the gradient that we use to train our model end-to-end with a normal SGD procedure. This new method is evaluated on three classification tasks, where we show that it allows to reduce computation and memory footprint by an order of magnitude for the same accuracy as classical architectures. We also show the consistency of the sampling that indeed focuses on informative parts of the input images.","url_abs":"https://arxiv.org/abs/1905.03711v2","url_pdf":"https://arxiv.org/pdf/1905.03711v2.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":"190503711","repo_url":"https://github.com/idiap/attention-sampling","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"ok","spdx":"NOASSERTION"}},{"paper_slug":"190503711","repo_url":"https://github.com/sara-nl/attention-sampling-pytorch","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"deep-attention","task_name":"Deep Attention"},{"task_slug":null,"task_name":"GPU"}],"methods":[{"method_slug":"sgd","method_name":"SGD"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1905.03711","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1905.03711"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+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/sara-nl/attention-sampling-pytorch","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/idiap/attention-sampling","reach":{"status":"ok","spdx":"NOASSERTION"}}],"summary":{"unverified":6},"by_repo_kind":{"listed":{"samples":6,"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":"34df8e42b75c38a9","entry":"batch_norm","repo":"sara-nl/attention-sampling-pytorch","repo_kind":"listed","path":"models/feature_model.py","file_url":"https://github.com/sara-nl/attention-sampling-pytorch/blob/HEAD/models/feature_model.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":"34df8e42b75c38a9"}},{"code_sha256_prefix":"e7efc21b9ad42500","entry":"calc_cls_measures","repo":"sara-nl/attention-sampling-pytorch","repo_kind":"listed","path":"utils.py","file_url":"https://github.com/sara-nl/attention-sampling-pytorch/blob/HEAD/utils.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":"e7efc21b9ad42500"}},{"code_sha256_prefix":"70d26e5a7c6ba2cf","entry":"conv_layer","repo":"sara-nl/attention-sampling-pytorch","repo_kind":"listed","path":"models/feature_model.py","file_url":"https://github.com/sara-nl/attention-sampling-pytorch/blob/HEAD/models/feature_model.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":"70d26e5a7c6ba2cf"}},{"code_sha256_prefix":"326f2acb728d830c","entry":"move_to","repo":"sara-nl/attention-sampling-pytorch","repo_kind":"listed","path":"utils.py","file_url":"https://github.com/sara-nl/attention-sampling-pytorch/blob/HEAD/utils.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":"326f2acb728d830c"}},{"code_sha256_prefix":"5891d1a7d387e2c5","entry":"sample","repo":"sara-nl/attention-sampling-pytorch","repo_kind":"listed","path":"ats/core/sampling.py","file_url":"https://github.com/sara-nl/attention-sampling-pytorch/blob/HEAD/ats/core/sampling.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":"5891d1a7d387e2c5"}},{"code_sha256_prefix":"76877081e7d2efbc","entry":"unravel_index","repo":"sara-nl/attention-sampling-pytorch","repo_kind":"listed","path":"ats/core/sampling.py","file_url":"https://github.com/sara-nl/attention-sampling-pytorch/blob/HEAD/ats/core/sampling.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":"76877081e7d2efbc"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}