Add Spectra-AASIST submission: CFAD

submissions/spectra-aasist.yaml

@@ -0,0 +1,38 @@
+schema_version: 4
+system:
+  name: Spectra-AASIST
+  slug: spectra-aasist
+  description: 'Speech deepfake detector: a wav2vec 2.0 XLS-R-300m self-supervised front-end feeding a
+    single-layer MLP bridge (1024->128) into an AASIST 2-class classifier with four-branch spectro-temporal
+    graph attention. Pre-release lab260/Spectra-AASIST checkpoint. FP32, preemphasis (0.97), deterministic
+    first-64600-sample window (no random crop). score = output logit for class 1 (bona fide). Unpublished
+    / pre-release model (no paper).
+
+    '
+  code: https://huggingface.co/lab260/Spectra-AASIST
+  checkpoint: https://huggingface.co/lab260/Spectra-AASIST
+  params_millions: 316.0122
+dataset:
+  id: SpeechAntiSpoofingBenchmarks/CFAD
+  revision: 53d7855c1c378524f7b7b1030bcb6b2caa327fe6
+  split: test
+scores:
+  eer_percent: 0.48095238095238096
+  n_trials: 62999
+  n_skipped: 0
+artifact:
+  scores_url: https://huggingface.co/lab260/Spectra-AASIST/resolve/8f1f2f4b947f2eacb4c77a887f89befd8ff70b05/.eval_results/SpeechAntiSpoofingBenchmarks/CFAD/scores.txt
+  scores_sha256: 09dd85ba7f5cc4097a7fccd0c4a335a5c3d87cb6e201eaa590440297512770a9
+  bench_version: speech-spoof-bench==0.3.4
+submitter:
+  hf_username: korallll
+  contact: k.n.borodin@mtuci.ru
+submitted_at: '2026-06-07'
+notes: Spectra-AASIST (wav2vec2 XLS-R-300m SSL front-end + MLP bridge + AASIST). Preemphasis 0.97, deterministic
+  first-64600-sample window. score = logit class 1 (bona fide). Unpublished/pre-release (no paper). Scores
+  hosted in lab260/Spectra-AASIST/.eval_results.
+reproduction:
+  reproduced_by: SpeechAntiSpoofingBenchmarks
+  reproduced_at: '2026-06-07'
+  reproduced_bench_version: speech-spoof-bench==0.3.4
+  match: scoring