{
  "$schema": "https://themachinepress.com/schemas/story-v1.schema.json",
  "schema_version": "1.0.0",
  "document_type": "machine_press_story",
  "story": {
    "story_id": "mp-2026-08-31-007",
    "source_story_id": "tmp-story-signrr-retrieve-refine",
    "edition_id": "mp-2026-08-31-morning-0053",
    "edition_url": "https://themachinepress.com/edition/2026-08-31",
    "position": 7,
    "story_type": "dispatch",
    "section": "research",
    "editorial_classification": "editorial",
    "headline": "Real Signing Motion Entered Before Generation",
    "slug": "real-signing-motion-entered-before-generation",
    "dek": "SignRR retrieves articulated sign segments and refines the full sequence instead of synthesizing motion from scratch.",
    "summary": "SignRR retrieves articulated sign segments and refines the full sequence instead of synthesizing motion from scratch.",
    "body_text": "Retrieval preserves rare hand configurations and signer articulation, but direct concatenation can break rhythm and style. SignRR begins with a dictionary of real motion segments, then uses a part-aware residual VQ-VAE to refine the complete sequence and handle length differences in latent space. On PHOENIX14T and CSL-Daily, the authors report state-of-the-art back-translation with competitive pose quality; the benchmarks do not establish fluency across all sign languages or signers.",
    "why_it_matters": "SignRR retrieves articulated sign segments and refines the full sequence instead of synthesizing motion from scratch.",
    "limitations": [
      "On PHOENIX14T and CSL-Daily, the authors report state-of-the-art back-translation with competitive pose quality; the benchmarks do not establish fluency across all sign languages or signers."
    ],
    "importance": 8,
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-31-007/real-signing-motion-entered-before-generation",
    "json_url": "https://themachinepress.com/story/mp-2026-08-31-007.json",
    "first_published_at": "2026-08-31T09:00:00.000-04:00",
    "modified_at": "2026-08-31T09:00:00.000-04:00",
    "content_status": "new",
    "is_carryover": false,
    "carryover_reason": null,
    "key_claims": [
      {
        "claim_id": "claim-mp-2026-08-31-007-001",
        "text": "SignRR retrieves articulated sign segments and refines the full sequence instead of synthesizing motion from scratch.",
        "source_ids": [
          "source-2026-08-31-007"
        ],
        "qualification": "On PHOENIX14T and CSL-Daily, the authors report state-of-the-art back-translation with competitive pose quality; the benchmarks do not establish fluency across all sign languages or signers."
      }
    ],
    "source_ids": [
      "source-2026-08-31-007"
    ],
    "tags": [
      "sign language",
      "motion generation",
      "accessibility"
    ],
    "image_url": null,
    "corrections": []
  },
  "sources": [
    {
      "source_id": "source-2026-08-31-007",
      "title": "arXiv preprint 2608.28568",
      "publisher": "arXiv",
      "url": "https://arxiv.org/abs/2608.28568",
      "canonical_url": "https://arxiv.org/abs/2608.28568",
      "source_type": "primary_research",
      "is_primary_source": true,
      "published_at": "2026-08-28T13:45:14.000-04:00",
      "accessed_at": "2026-08-31T08:28:26.000-04:00",
      "supports_claim_ids": [
        "claim-mp-2026-08-31-007-001"
      ]
    }
  ],
  "corrections": [],
  "publisher": {
    "name": "The Machine Press",
    "url": "https://themachinepress.com",
    "description": "A daily newspaper for the age of artificial intelligence."
  },
  "cite_this_report": {
    "title": "Real Signing Motion Entered Before Generation",
    "publisher": "The Machine Press",
    "published_at": "2026-08-31T09:00:00.000-04:00",
    "canonical_url": "https://themachinepress.com/story/mp-2026-08-31-007/real-signing-motion-entered-before-generation"
  }
}