{
  "reviewed_utc": "2026-09-19T04:13:24.593439+00:00",
  "stage": "first novelty review after entry018 pilot",
  "claim_under_review": "A controlled demonstration that a bird classifier loses target detections when another bird occupies a disjoint part of the same window, and the resulting artificial association between independently present species.",
  "queries": [
    "\"bird\" \"classifier\" \"non-overlapping\" calls",
    "\"acoustic\" \"co-occurrence\" \"classifier\" bias",
    "\"BirdNET\" \"suppression\" calls",
    "\"BirdNET\" \"temporal\" \"masking\"",
    "\"bird\" \"classification\" \"temporal context\" interference",
    "\"BirdNET\" \"mixing\" \"confidence\"",
    "\"bioacoustic\" \"spurious associations\"",
    "\"bird\" \"classifier\" \"co-detection\"",
    "\"Improving bird classification with unsupervised sound separation\"",
    "\"imperfect detection\" \"spurious\" \"co-occurrence\"",
    "\"acoustic\" \"species interactions\" \"detection probability\" classifier",
    "\"bird\" \"co-occurrence\" \"classification errors\"",
    "\"BirdNET\" \"nonoverlapping\" mixtures"
  ],
  "nearest_prior": [
    {
      "title": "Clark et al. 2023, soundscape composition",
      "url": "https://doi.org/10.1016/j.ecoinf.2023.102065",
      "scope": "Observed soundscape classes and errors; broad biophony class did not resolve interference among animal vocalizations. Uses sigmoid multi-label outputs and discusses overlapping sounds.",
      "overlap": "General soundscape-dependent detection is prior knowledge; no claim to discover it."
    },
    {
      "title": "Denton, Wisdom and Hershey, ICASSP2022",
      "url": "https://arxiv.org/abs/2110.03209",
      "scope": "Unsupervised separation of overlapping songs/noise improves classification; pooling original and separated channels is prior art.",
      "overlap": "Source separation and max-pooling are not novel proposed methods."
    },
    {
      "title": "Sasek et al. 2024",
      "url": "https://doi.org/10.7717/peerj.17854",
      "scope": "Synthetic overlapping background/target mixtures; site-specific source separation and BirdNET robustness tests for Golden-cheeked Warbler.",
      "overlap": "Synthetic mixture interventions and classifier robustness are established."
    },
    {
      "title": "Separating Overlapping Birdsongs Enhances the Reliability of Avian Vocal Activity Analysis, 2026",
      "url": "https://doi.org/10.1002/ece3.73648",
      "scope": "30-species synthetic overlapping mixtures, source separation and ecological vocal-activity analyses; full text examined.",
      "overlap": "Ecological consequences of overlapping-source classification errors are already studied."
    },
    {
      "title": "Multi-species Mixing for Weakly Supervised SED Under Domain Shift",
      "url": "https://www.dfki.de/en/web/research/projects-and-publications/publication/17345",
      "scope": "Synthetic multi-species training augmentation with BirdNET embeddings and linear classification, evaluated on five anurans.",
      "overlap": "Temporal mixture augmentation is established."
    },
    {
      "title": "How to analyse overlapping sounds in the marine environment using supervised multi-label classification, 2026",
      "url": "https://doi.org/10.1038/s44384-026-00060-x",
      "scope": "Naturally co-occurring marine sounds; training co-occurrence structure and architectural interference.",
      "overlap": "Classifier dependence on other sound classes is known across bioacoustics."
    }
  ],
  "assessment": "No exact controlled disjoint-time bird counterfactual with a measured artificial species-association result was located in these searches. This is a provisional narrow gap, not proof of novelty. Proceed to held-out recording validation, then repeat novelty review.",
  "preprocessing_audit": "Pinned BirdNET v2.4 TFLite graph begins with waveform min/max normalization, and later uses global pooling. A second source may change model preprocessing even when raw target samples remain identical. Mechanism not yet attributed.",
  "followup_prior_before_focal_outcomes": [
    {
      "title": "Richmond, Hines and Beissinger2010",
      "url": "https://doi.org/10.1890/09-0470.1",
      "scope": "Conditional two-species occupancy models already account for false absences and detection dependence. Do not claim a new ecological statistical principle."
    },
    {
      "title": "Tobler et al.2019",
      "url": "https://doi.org/10.1002/ecy.2754",
      "scope": "JSDMs with imperfect detection; detection heterogeneity can distort inferred species correlations."
    },
    {
      "title": "Metcalf et al.2022",
      "url": "https://doi.org/10.1111/2041-210X.13967",
      "scope": "Acoustic classification error heterogeneity and contextual correction; recommends validation for ecological predictor analyses."
    },
    {
      "title": "Chambert et al.2018",
      "url": "https://doi.org/10.1111/2041-210X.12985",
      "scope": "Two-species occupancy models accommodating misidentification and non-detection."
    }
  ]
}