Add XLSR-Mamba-DF submission for LRLspoof (1-SRR=24.3183)

submissions/xlsr-mamba-df.yaml

@@ -0,0 +1,52 @@
+schema_version: 4
+system:
+  name: XLSR-Mamba-DF
+  slug: xlsr-mamba-df
+  description: 'XLSR-Mamba (DF checkpoint): a fully fine-tuned wav2vec 2.0 XLS-R 300M front-end (last
+    layer) projected to 144-d (Linear + BatchNorm + SELU), followed by a dual-column bidirectional Mamba
+    (v1 selective SSM) back-end: 6 forward + 6 backward Mamba blocks with fused add+RMSNorm, per-direction
+    attention pooling, concatenation, linear fusion and a 2-way head. Official authors'' DF-track checkpoint
+    (trained on ASVspoof2019 LA train with RawBoost algo 3). Not to be confused with XLSR-MamBo (arXiv
+    2601.02944), a different model.'
+  code: https://github.com/swagshaw/XLSR-Mamba
+  checkpoint: https://huggingface.co/SpeechAntiSpoofingBenchmarks/XLSR-Mamba-DF/blob/e97533f5ac8303cf0e7215bb8b452cdc7a74fefb/model.safetensors
+  params_millions: 319.3275
+  paper:
+    arxiv_id: '2411.10027'
+    url: https://arxiv.org/abs/2411.10027
+    bibtex: "@article{xiao2024xlsr,\n  title   = {{XLSR-Mamba}: A Dual-Column Bidirectional State Space\
+      \ Model for Spoofing Attack Detection},\n  author  = {Xiao, Yang and Das, Rohan Kumar},\n  journal\
+      \ = {IEEE Signal Processing Letters},\n  note    = {arXiv:2411.10027},\n  year    = {2025}\n}\n"
+dataset:
+  id: lab260/LRLspoof
+  revision: ac76f0e5472fdb68a6726e1fa94a148316e76916
+  split: test
+scores:
+  srr_complement: 24.31832070843579
+  n_trials: 1304455
+  n_skipped: 286
+calibration:
+  source_dataset: SpeechAntiSpoofingBenchmarks/DeepVoice
+  threshold: -1.715145132165605
+artifact:
+  scores_url: https://huggingface.co/SpeechAntiSpoofingBenchmarks/XLSR-Mamba-DF/resolve/ebe3647bd61f294812853355f74e01ec7af9f808/.eval_results/lab260/LRLspoof/scores.txt
+  scores_sha256: d74a21741ed47e21f32c7ca02e101d2dbc135a0222e3dc19f8b3060490ad8eac
+  bench_version: speech-spoof-bench==0.4.1
+reproduction:
+  reproduced_by: SpeechAntiSpoofingBenchmarks
+  reproduced_at: '2026-11-01'
+  reproduced_bench_version: speech-spoof-bench==0.4.1
+  match: inference
+submitter:
+  hf_username: korallll
+  contact: k.n.borodin@mtuci.ru
+submitted_at: '2026-11-01'
+notes: 'XLSR-Mamba (Xiao & Das, IEEE SPL; arXiv 2411.10027), MIT, checkpoint HF AustinXiao/XLSR-Mamba-DF
+  model.safetensors @ 8aed3da9242b161f63647b099f67b3c648acbd00, strict load (319,327,542 parameters +
+  3 BN buffer elements). Deterministic first-66800-sample window (4.175 s), shorter clips repeat-tiled
+  (source utils.pad); no resampling or normalisation; FP32; bona-fide score = logits[:, 1] (source main.py).
+  Deviations: Mamba selective scan, causal depthwise conv and RMSNorm in plain PyTorch (same math as the
+  mamba_ssm CUDA/Triton kernels; chunked scan == selective_scan_ref within 1e-4) on every GPU -- mamba_ssm/causal_conv1d
+  are not installed and Triton is unreliable on sm_120; the source''s duplicate (unused) XLS-R forward
+  and its extract_feat train() call are removed. Not to be confused with XLSR-MamBo (arXiv 2601.02944,
+  a different model).'