{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2020:SEG5KOYF322JWWSXXTB2EQBOIZ","short_pith_number":"pith:SEG5KOYF","schema_version":"1.0","canonical_sha256":"910dd53b05deb49b5a57bcc3a2402e4645b6be2e0d12acdf597ea243c0ce64eb","source":{"kind":"arxiv","id":"2004.13649","version":4},"attestation_state":"computed","paper":{"title":"Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","eess.IV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Denis Yarats, Ilya Kostrikov, Rob Fergus","submitted_at":"2020-04-28T16:48:16Z","abstract_excerpt":"We propose a simple data augmentation technique that can be applied to standard model-free reinforcement learning algorithms, enabling robust learning directly from pixels without the need for auxiliary losses or pre-training. The approach leverages input perturbations commonly used in computer vision tasks to regularize the value function. Existing model-free approaches, such as Soft Actor-Critic (SAC), are not able to train deep networks effectively from image pixels. However, the addition of our augmentation method dramatically improves SAC's performance, enabling it to reach state-of-the-a"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2004.13649","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2020-04-28T16:48:16Z","cross_cats_sorted":["cs.CV","eess.IV","stat.ML"],"title_canon_sha256":"db434fd6b3852b7cd6fd1b4841377bc6b4eb1e29f46701308d29afd2b71d41f3","abstract_canon_sha256":"9cb3f9dc70d8212ff73c9b9d23a38c606b2e1c3a2077eea854f2b14b1614cb67"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:20:54.357863Z","signature_b64":"XxQR0erECCusCeQNikMF05o8pUrU1/VAqNytx3KllihvDeVsjMNLBE4az9ubGpcbMeWiBtBAu2x0oM0I58v+Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"910dd53b05deb49b5a57bcc3a2402e4645b6be2e0d12acdf597ea243c0ce64eb","last_reissued_at":"2026-07-05T02:20:54.357278Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:20:54.357278Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixels","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CV","eess.IV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Denis Yarats, Ilya Kostrikov, Rob Fergus","submitted_at":"2020-04-28T16:48:16Z","abstract_excerpt":"We propose a simple data augmentation technique that can be applied to standard model-free reinforcement learning algorithms, enabling robust learning directly from pixels without the need for auxiliary losses or pre-training. The approach leverages input perturbations commonly used in computer vision tasks to regularize the value function. Existing model-free approaches, such as Soft Actor-Critic (SAC), are not able to train deep networks effectively from image pixels. However, the addition of our augmentation method dramatically improves SAC's performance, enabling it to reach state-of-the-a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.13649","kind":"arxiv","version":4},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2004.13649/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2004.13649","created_at":"2026-07-05T02:20:54.357359+00:00"},{"alias_kind":"arxiv_version","alias_value":"2004.13649v4","created_at":"2026-07-05T02:20:54.357359+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.13649","created_at":"2026-07-05T02:20:54.357359+00:00"},{"alias_kind":"pith_short_12","alias_value":"SEG5KOYF322J","created_at":"2026-07-05T02:20:54.357359+00:00"},{"alias_kind":"pith_short_16","alias_value":"SEG5KOYF322JWWSX","created_at":"2026-07-05T02:20:54.357359+00:00"},{"alias_kind":"pith_short_8","alias_value":"SEG5KOYF","created_at":"2026-07-05T02:20:54.357359+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":11,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.24078","citing_title":"MinInter: Minimizing Trajectory Interpolation During Data Augmentation for Imitation Learning","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20104","citing_title":"Sensorimotor World Models: Perception for Action via Inverse Dynamics","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03800","citing_title":"Trading Human Curation for Synthetic Augmentation in RLVR","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2510.17640","citing_title":"RESample: A Robust Data Augmentation Framework via Exploratory Sampling for Robotic Manipulation","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2302.11550","citing_title":"Scaling Robot Learning with Semantically Imagined Experience","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2310.17596","citing_title":"MimicGen: A Data Generation System for Scalable Robot Learning using Human Demonstrations","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2203.12601","citing_title":"R3M: A Universal Visual Representation for Robot Manipulation","ref_index":27,"is_internal_anchor":false},{"citing_arxiv_id":"2010.02193","citing_title":"Mastering Atari with Discrete World Models","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2108.03298","citing_title":"What Matters in Learning from Offline Human Demonstrations for Robot Manipulation","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24661","citing_title":"Agent-Centric Observation Adaptation for Robust Visual Control under Dynamic Perturbations","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24661","citing_title":"Agent-Centric Observation Adaptation for Robust Visual Control under Dynamic Perturbations","ref_index":30,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ","json":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ.json","graph_json":"https://pith.science/api/pith-number/SEG5KOYF322JWWSXXTB2EQBOIZ/graph.json","events_json":"https://pith.science/api/pith-number/SEG5KOYF322JWWSXXTB2EQBOIZ/events.json","paper":"https://pith.science/paper/SEG5KOYF"},"agent_actions":{"view_html":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ","download_json":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ.json","view_paper":"https://pith.science/paper/SEG5KOYF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2004.13649&json=true","fetch_graph":"https://pith.science/api/pith-number/SEG5KOYF322JWWSXXTB2EQBOIZ/graph.json","fetch_events":"https://pith.science/api/pith-number/SEG5KOYF322JWWSXXTB2EQBOIZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ/action/storage_attestation","attest_author":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ/action/author_attestation","sign_citation":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ/action/citation_signature","submit_replication":"https://pith.science/pith/SEG5KOYF322JWWSXXTB2EQBOIZ/action/replication_record"}},"created_at":"2026-07-05T02:20:54.357359+00:00","updated_at":"2026-07-05T02:20:54.357359+00:00"}