Add RawTFNet scores for DECRO

#12
by korallll - opened
Files changed (1) hide show
  1. submissions/rawtfnet.yaml +37 -0
submissions/rawtfnet.yaml ADDED
@@ -0,0 +1,37 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ schema_version: 4
2
+ system:
3
+ name: RawTFNet
4
+ slug: rawtfnet
5
+ description: 'Lightweight raw-waveform CNN for speech anti-spoofing: sinc-convolution front-end, depthwise-separable Res2Net-SE
6
+ conv blocks, and a Tf-SepNet (time-frequency separable) classifier (depth=10, width=32). ASVspoof2019 LA pretrained, FP32,
7
+ deterministic first-64600-sample window (no random crop).'
8
+ code: https://github.com/swagshaw/RawTFNet-Pytorch
9
+ checkpoint: https://huggingface.co/SpeechAntiSpoofingBenchmarks/RawTFNet/blob/aa12f0fe2f10cc5278c954175a12c18cc43e3113/Best_RawTFNet_32.pth
10
+ params_millions: 0.17754
11
+ paper:
12
+ arxiv_id: '2507.08227'
13
+ url: https://arxiv.org/abs/2507.08227
14
+ bibtex: "@article{xiao2025rawtfnet,\n title={RawTFNet: A Lightweight CNN Architecture for Speech Anti-spoofing},\n author={Xiao,\
15
+ \ Yang and Dang, Ting and Das, Rohan Kumar},\n journal={arXiv preprint arXiv:2507.08227},\n year={2025}\n}\n"
16
+ dataset:
17
+ id: SpeechAntiSpoofingBenchmarks/DECRO
18
+ revision: bb6df2524eadaab6aa6a2366a41a2a5fe1e4104d
19
+ split: test
20
+ scores:
21
+ eer_percent: 23.72578906279044
22
+ n_trials: 37314
23
+ n_skipped: 0
24
+ artifact:
25
+ scores_url: https://huggingface.co/SpeechAntiSpoofingBenchmarks/RawTFNet/resolve/09139ca715c6a44e9bd5a7ae752c0688894a9e28/.eval_results/SpeechAntiSpoofingBenchmarks/DECRO/scores.txt
26
+ scores_sha256: a7e2f8d906b7780f8cca37c5b32054c11c8ebd789f8711d9810e4d55d53aae13
27
+ bench_version: speech-spoof-bench==0.3.4
28
+ reproduction:
29
+ reproduced_by: SpeechAntiSpoofingBenchmarks
30
+ reproduced_at: '2026-06-10'
31
+ reproduced_bench_version: speech-spoof-bench==0.3.4
32
+ match: scoring
33
+ submitter:
34
+ hf_username: korallll
35
+ contact: k.n.borodin@mtuci.ru
36
+ submitted_at: '2026-06-10'
37
+ notes: Deterministic first-64600-sample window (no random crop). Full RawTFNet (width=32), Best_RawTFNet_32.pth.