Add random-baseline submission (EER 49.56%)

submissions/random-baseline.yaml

@@ -0,0 +1,64 @@
+schema_version: 4
+
+system:
+  name: random-baseline
+  slug: random-baseline
+  description: >
+    Reference random baseline. Returns N(0, 1) for every utterance using a
+    fixed seed (seed=0). EER ≈ 50% by construction. Used as the seeded
+    smoke-test baseline for the arena (roadmap Phase 3).
+  code: https://github.com/SpeechAntiSpoofingBenchmarks/speech-spoof-bench
+  checkpoint: https://huggingface.co/SpeechAntiSpoofingBenchmarks/random-baseline-asas
+  params_millions: 1
+  paper:
+    arxiv_id: "1911.01601"
+    url: https://arxiv.org/abs/1911.01601
+    bibtex: |
+      @article{wang2020asvspoof,
+        title={ASVspoof 2019: A large-scale public database of synthesized,
+               converted and replayed speech},
+        author={Wang, Xin and Yamagishi, Junichi and Todisco, Massimiliano and
+                Delgado, H{\'e}ctor and Nautsch, Andreas and Evans, Nicholas
+                and Sahidullah, Md and Vestman, Ville and Kinnunen, Tomi and
+                Lee, Kong Aik and others},
+        journal={Computer Speech \& Language},
+        volume={64},
+        pages={101114},
+        year={2020},
+        publisher={Elsevier}
+      }
+
+dataset:
+  id: SpeechAntiSpoofingBenchmarks/SONAR
+  revision: bf65a137339acbe91f4baa702b7638fcda69aca3
+  split: test
+
+scores:
+  eer_percent: 49.56024626209323
+  n_trials: 4548
+  n_skipped: 0
+
+artifact:
+  scores_url: https://huggingface.co/SpeechAntiSpoofingBenchmarks/random-baseline-asas/resolve/8af0ccc677cd89e67f6bdfd4514bc715137fc0dc/.eval_results/SpeechAntiSpoofingBenchmarks/SONAR/scores.txt
+  scores_sha256: 051cc20027b196b4b8a69986d7c17017dac5f0ee2c69138198ea6cd65f70dfe7
+  bench_version: speech-spoof-bench==0.3.4
+
+reproduction:
+  reproduced_by: SpeechAntiSpoofingBenchmarks
+  reproduced_at: 2026-06-07
+  reproduced_bench_version: speech-spoof-bench==0.3.4
+  match: scoring
+
+submitter:
+  hf_username: SpeechAntiSpoofingBenchmarks
+  contact: k.n.borodin@mtuci.ru
+
+submitted_at: 2026-06-07
+
+notes: >
+  Seeded random baseline for the newly-added SONAR dataset. Random baseline has
+  no associated paper of its own; the paper field cites the ASVspoof 2019 dataset
+  paper as a placeholder, consistent with the other random-baseline submissions
+  (same system slug). The reproduction block is left empty for the maintainer to
+  fill at merge. verify-pr only routes once SONAR is ingested (manifest PR merged
+  + re-ingest), so the verify/merge half waits on the maintainer.