{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:M5JGU57VEH7A7QE5F4EZM2QPES","short_pith_number":"pith:M5JGU57V","schema_version":"1.0","canonical_sha256":"67526a77f521fe0fc09d2f09966a0f24976c6b984c31c68d08a388bda13f3a46","source":{"kind":"arxiv","id":"1912.04958","version":2},"attestation_state":"computed","paper":{"title":"Analyzing and Improving the Image Quality of StyleGAN","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.NE","eess.IV","stat.ML"],"primary_cat":"cs.CV","authors_text":"Jaakko Lehtinen, Janne Hellsten, Miika Aittala, Samuli Laine, Tero Karras, Timo Aila","submitted_at":"2019-12-03T11:44:01Z","abstract_excerpt":"The style-based GAN architecture (StyleGAN) yields state-of-the-art results in data-driven unconditional generative image modeling. We expose and analyze several of its characteristic artifacts, and propose changes in both model architecture and training methods to address them. In particular, we redesign the generator normalization, revisit progressive growing, and regularize the generator to encourage good conditioning in the mapping from latent codes to images. In addition to improving image quality, this path length regularizer yields the additional benefit that the generator becomes signi"},"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":"1912.04958","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2019-12-03T11:44:01Z","cross_cats_sorted":["cs.LG","cs.NE","eess.IV","stat.ML"],"title_canon_sha256":"6f17c44f5e5e0c7e8abf51fcb51c81a13f76183976e705f301cca68e1775f012","abstract_canon_sha256":"df1f96b2affea3c2b556e57c914de2c9465c998488a8e98c431544d59d5268ae"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:49:53.033186Z","signature_b64":"mAV4sJTLURppLadWI06WGBizwbG2zviNtpoAUkbKVKrSbsQ76I6AxTAsGeOBdqHpI1dUaDR6QqGm2iCwvBsBCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"67526a77f521fe0fc09d2f09966a0f24976c6b984c31c68d08a388bda13f3a46","last_reissued_at":"2026-07-05T00:49:53.032698Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:49:53.032698Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analyzing and Improving the Image Quality of StyleGAN","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG","cs.NE","eess.IV","stat.ML"],"primary_cat":"cs.CV","authors_text":"Jaakko Lehtinen, Janne Hellsten, Miika Aittala, Samuli Laine, Tero Karras, Timo Aila","submitted_at":"2019-12-03T11:44:01Z","abstract_excerpt":"The style-based GAN architecture (StyleGAN) yields state-of-the-art results in data-driven unconditional generative image modeling. We expose and analyze several of its characteristic artifacts, and propose changes in both model architecture and training methods to address them. In particular, we redesign the generator normalization, revisit progressive growing, and regularize the generator to encourage good conditioning in the mapping from latent codes to images. In addition to improving image quality, this path length regularizer yields the additional benefit that the generator becomes signi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1912.04958","kind":"arxiv","version":2},"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/1912.04958/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":"1912.04958","created_at":"2026-07-05T00:49:53.032759+00:00"},{"alias_kind":"arxiv_version","alias_value":"1912.04958v2","created_at":"2026-07-05T00:49:53.032759+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1912.04958","created_at":"2026-07-05T00:49:53.032759+00:00"},{"alias_kind":"pith_short_12","alias_value":"M5JGU57VEH7A","created_at":"2026-07-05T00:49:53.032759+00:00"},{"alias_kind":"pith_short_16","alias_value":"M5JGU57VEH7A7QE5","created_at":"2026-07-05T00:49:53.032759+00:00"},{"alias_kind":"pith_short_8","alias_value":"M5JGU57V","created_at":"2026-07-05T00:49:53.032759+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":9,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05912","citing_title":"Self-Learning Expression Deformations for Data-Efficient Gaussian Avatars","ref_index":34,"is_internal_anchor":false},{"citing_arxiv_id":"2606.10023","citing_title":"Learning the Universe: Posterior Reliability of Neural Generative Models in High-Dimensional Field-Level Inference of Cosmic Initial Conditions","ref_index":62,"is_internal_anchor":false},{"citing_arxiv_id":"1912.08786","citing_title":"Why we need an AI-resilient society","ref_index":12,"is_internal_anchor":false},{"citing_arxiv_id":"2101.02388","citing_title":"Knowledge Distillation in Iterative Generative Models for Improved Sampling Speed","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2105.05233","citing_title":"Diffusion Models Beat GANs on Image Synthesis","ref_index":28,"is_internal_anchor":false},{"citing_arxiv_id":"2112.10752","citing_title":"High-Resolution Image Synthesis with Latent Diffusion Models","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07724","citing_title":"Curated Synthetic Data Doesn't Have to Collapse: A Theoretical Study of Generative Retraining with Pluralistic Preferences","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06678","citing_title":"A Wasserstein GAN-based climate scenario generator for risk management and insurance: the case of soil subsidence","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20936","citing_title":"AttentionBender: Manipulating Cross-Attention in Video Diffusion Transformers as a Creative Probe","ref_index":37,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES","json":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES.json","graph_json":"https://pith.science/api/pith-number/M5JGU57VEH7A7QE5F4EZM2QPES/graph.json","events_json":"https://pith.science/api/pith-number/M5JGU57VEH7A7QE5F4EZM2QPES/events.json","paper":"https://pith.science/paper/M5JGU57V"},"agent_actions":{"view_html":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES","download_json":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES.json","view_paper":"https://pith.science/paper/M5JGU57V","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1912.04958&json=true","fetch_graph":"https://pith.science/api/pith-number/M5JGU57VEH7A7QE5F4EZM2QPES/graph.json","fetch_events":"https://pith.science/api/pith-number/M5JGU57VEH7A7QE5F4EZM2QPES/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES/action/timestamp_anchor","attest_storage":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES/action/storage_attestation","attest_author":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES/action/author_attestation","sign_citation":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES/action/citation_signature","submit_replication":"https://pith.science/pith/M5JGU57VEH7A7QE5F4EZM2QPES/action/replication_record"}},"created_at":"2026-07-05T00:49:53.032759+00:00","updated_at":"2026-07-05T00:49:53.032759+00:00"}