Hugging Face's logo

niobures
/
FullSubNet

Model card Files
xet
Community
niobures commited on
Commit
0d87a24
·
verified ·
1 Parent(s): 06124be

FullSubNet, FullSubNet+, Fast-FullSubNet, Mel-FullSubNet, Spiking-FullSubNet

Browse files
Files changed (36) hide show
  1. .gitattributes +11 -0
  2. Fast-FullSubNet/Fast FullSubNet. Accelerate Full-band and Sub-band Fusion Model for Single-channel Speech Enhancement.pdf +3 -0
  3. Fast-FullSubNet/models/Fast-FullSubNet/.gitattributes +35 -0
  4. Fast-FullSubNet/models/Fast-FullSubNet/README.md +36 -0
  5. Fast-FullSubNet/models/Fast-FullSubNet/fast_fullsubnet_best_model_118epochs.tar +3 -0
  6. Fast-FullSubNet/models/Fast-FullSubNet/source.txt +1 -0
  7. FullSubNet+/FullSubNet+. Channel Attention FullSubNet with Complex Spectrograms for Speech Enhancement.pdf +3 -0
  8. FullSubNet+/Learnable spectral dimension compression mapping for full-band speech enhancement.pdf +3 -0
  9. FullSubNet+/code/FullSubNet-plus [freds0] +2 -3.zip +3 -0
  10. FullSubNet+/code/FullSubNet-plus-optimizations.zip +3 -0
  11. FullSubNet+/code/FullSubNet-plus.zip +3 -0
  12. FullSubNet+/models/best_model.tar +3 -0
  13. FullSubNet/FullSubNet. A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement.pdf +3 -0
  14. FullSubNet/Improving the Speech Enhancement Model with Discrete Wavelet Transform Sub-Band Features in Adaptive FullSubNet.pdf +3 -0
  15. FullSubNet/Speech Enhancement with Fullband-Subband Cross-Attention Network.pdf +3 -0
  16. FullSubNet/code/FullSubNet (original)/FullSubNet.zip +3 -0
  17. FullSubNet/code/FullSubNet (original)/v0.2/Checkpoints.txt +23 -0
  18. FullSubNet/code/FullSubNet (original)/v0.2/FullSubNet-0.2.zip +3 -0
  19. FullSubNet/code/FullSubNet (original)/v0.2/RIR.Multichannel.Impulse.Response.Database.+.The.REVERB.challenge.zip +3 -0
  20. FullSubNet/code/FullSubNet (original)/v0.2/Room Impulse Responses.txt +8 -0
  21. FullSubNet/code/FullSubNet (original)/v0.2/cum_fullsubnet_best_model_218epochs.tar +3 -0
  22. FullSubNet/code/FullSubNet (original)/v0.2/fullsubnet_best_model_58epochs.tar +3 -0
  23. FullSubNet/code/FullSubNet (original)/wiki.zip +3 -0
  24. FullSubNet/code/FullSubNetWithASR.zip +3 -0
  25. FullSubNet/code/SE-FullSubNet.zip +3 -0
  26. FullSubNet/code/fullsubnet_training.ipynb +0 -0
  27. Mel-FullSubNet/A Mel Spectrogram Enhancement Paradigm Based on CWT in Speech Synthesis.pdf +3 -0
  28. Mel-FullSubNet/Mel-FullSubNet. Mel-Spectrogram Enhancement for Improving Both Speech Quality and ASR.pdf +3 -0
  29. Spiking-FullSubNet/DPSNN_Spiking_Neural_Network_for_Low-Latency_Strea.pdf +3 -0
  30. Spiking-FullSubNet/Documentation.txt +1 -0
  31. Spiking-FullSubNet/Spiking-FullSubNet.pdf +3 -0
  32. Spiking-FullSubNet/Towards Ultra-Low-Power Neuromorphic Speech Enhancement with Spiking-FullSubNet.pdf +3 -0
  33. Spiking-FullSubNet/code/spiking-fullsubnet-inference.zip +3 -0
  34. Spiking-FullSubNet/code/spiking-fullsubnet.zip +3 -0
  35. Spiking-FullSubNet/data/spiking-fullsubnet-data.zip +3 -0
  36. Spiking-FullSubNet/data/validation_set.tar.gz +3 -0
.gitattributes CHANGED
@@ -33,3 +33,14 @@ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
 
 
 
 
 
 
 
 
 
 
 
 
33
  *.zip filter=lfs diff=lfs merge=lfs -text
34
  *.zst filter=lfs diff=lfs merge=lfs -text
35
  *tfevents* filter=lfs diff=lfs merge=lfs -text
36
+ Fast-FullSubNet/Fast[[:space:]]FullSubNet.[[:space:]]Accelerate[[:space:]]Full-band[[:space:]]and[[:space:]]Sub-band[[:space:]]Fusion[[:space:]]Model[[:space:]]for[[:space:]]Single-channel[[:space:]]Speech[[:space:]]Enhancement.pdf filter=lfs diff=lfs merge=lfs -text
37
+ FullSubNet/FullSubNet.[[:space:]]A[[:space:]]Full-Band[[:space:]]and[[:space:]]Sub-Band[[:space:]]Fusion[[:space:]]Model[[:space:]]for[[:space:]]Real-Time[[:space:]]Single-Channel[[:space:]]Speech[[:space:]]Enhancement.pdf filter=lfs diff=lfs merge=lfs -text
38
+ FullSubNet/Improving[[:space:]]the[[:space:]]Speech[[:space:]]Enhancement[[:space:]]Model[[:space:]]with[[:space:]]Discrete[[:space:]]Wavelet[[:space:]]Transform[[:space:]]Sub-Band[[:space:]]Features[[:space:]]in[[:space:]]Adaptive[[:space:]]FullSubNet.pdf filter=lfs diff=lfs merge=lfs -text
39
+ FullSubNet/Speech[[:space:]]Enhancement[[:space:]]with[[:space:]]Fullband-Subband[[:space:]]Cross-Attention[[:space:]]Network.pdf filter=lfs diff=lfs merge=lfs -text
40
+ FullSubNet+/FullSubNet+.[[:space:]]Channel[[:space:]]Attention[[:space:]]FullSubNet[[:space:]]with[[:space:]]Complex[[:space:]]Spectrograms[[:space:]]for[[:space:]]Speech[[:space:]]Enhancement.pdf filter=lfs diff=lfs merge=lfs -text
41
+ FullSubNet+/Learnable[[:space:]]spectral[[:space:]]dimension[[:space:]]compression[[:space:]]mapping[[:space:]]for[[:space:]]full-band[[:space:]]speech[[:space:]]enhancement.pdf filter=lfs diff=lfs merge=lfs -text
42
+ Mel-FullSubNet/A[[:space:]]Mel[[:space:]]Spectrogram[[:space:]]Enhancement[[:space:]]Paradigm[[:space:]]Based[[:space:]]on[[:space:]]CWT[[:space:]]in[[:space:]]Speech[[:space:]]Synthesis.pdf filter=lfs diff=lfs merge=lfs -text
43
+ Mel-FullSubNet/Mel-FullSubNet.[[:space:]]Mel-Spectrogram[[:space:]]Enhancement[[:space:]]for[[:space:]]Improving[[:space:]]Both[[:space:]]Speech[[:space:]]Quality[[:space:]]and[[:space:]]ASR.pdf filter=lfs diff=lfs merge=lfs -text
44
+ Spiking-FullSubNet/DPSNN_Spiking_Neural_Network_for_Low-Latency_Strea.pdf filter=lfs diff=lfs merge=lfs -text
45
+ Spiking-FullSubNet/Spiking-FullSubNet.pdf filter=lfs diff=lfs merge=lfs -text
46
+ Spiking-FullSubNet/Towards[[:space:]]Ultra-Low-Power[[:space:]]Neuromorphic[[:space:]]Speech[[:space:]]Enhancement[[:space:]]with[[:space:]]Spiking-FullSubNet.pdf filter=lfs diff=lfs merge=lfs -text
Fast-FullSubNet/Fast FullSubNet. Accelerate Full-band and Sub-band Fusion Model for Single-channel Speech Enhancement.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1b650a8db2abd5f0c911343c8afe501d44f46932d5cb02e08c46183e8835ad99
3
+ size 268074
Fast-FullSubNet/models/Fast-FullSubNet/.gitattributes ADDED
@@ -0,0 +1,35 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ *.7z filter=lfs diff=lfs merge=lfs -text
2
+ *.arrow filter=lfs diff=lfs merge=lfs -text
3
+ *.bin filter=lfs diff=lfs merge=lfs -text
4
+ *.bz2 filter=lfs diff=lfs merge=lfs -text
5
+ *.ckpt filter=lfs diff=lfs merge=lfs -text
6
+ *.ftz filter=lfs diff=lfs merge=lfs -text
7
+ *.gz filter=lfs diff=lfs merge=lfs -text
8
+ *.h5 filter=lfs diff=lfs merge=lfs -text
9
+ *.joblib filter=lfs diff=lfs merge=lfs -text
10
+ *.lfs.* filter=lfs diff=lfs merge=lfs -text
11
+ *.mlmodel filter=lfs diff=lfs merge=lfs -text
12
+ *.model filter=lfs diff=lfs merge=lfs -text
13
+ *.msgpack filter=lfs diff=lfs merge=lfs -text
14
+ *.npy filter=lfs diff=lfs merge=lfs -text
15
+ *.npz filter=lfs diff=lfs merge=lfs -text
16
+ *.onnx filter=lfs diff=lfs merge=lfs -text
17
+ *.ot filter=lfs diff=lfs merge=lfs -text
18
+ *.parquet filter=lfs diff=lfs merge=lfs -text
19
+ *.pb filter=lfs diff=lfs merge=lfs -text
20
+ *.pickle filter=lfs diff=lfs merge=lfs -text
21
+ *.pkl filter=lfs diff=lfs merge=lfs -text
22
+ *.pt filter=lfs diff=lfs merge=lfs -text
23
+ *.pth filter=lfs diff=lfs merge=lfs -text
24
+ *.rar filter=lfs diff=lfs merge=lfs -text
25
+ *.safetensors filter=lfs diff=lfs merge=lfs -text
26
+ saved_model/**/* filter=lfs diff=lfs merge=lfs -text
27
+ *.tar.* filter=lfs diff=lfs merge=lfs -text
28
+ *.tar filter=lfs diff=lfs merge=lfs -text
29
+ *.tflite filter=lfs diff=lfs merge=lfs -text
30
+ *.tgz filter=lfs diff=lfs merge=lfs -text
31
+ *.wasm filter=lfs diff=lfs merge=lfs -text
32
+ *.xz filter=lfs diff=lfs merge=lfs -text
33
+ *.zip filter=lfs diff=lfs merge=lfs -text
34
+ *.zst filter=lfs diff=lfs merge=lfs -text
35
+ *tfevents* filter=lfs diff=lfs merge=lfs -text
Fast-FullSubNet/models/Fast-FullSubNet/README.md ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ---
2
+ license: mit
3
+ pipeline_tag: audio-to-audio
4
+ tags:
5
+ - denoising
6
+ - speech enhancement
7
+ - speech separation
8
+ - noise suppression
9
+ - realtime
10
+ ---
11
+
12
+ This is a pre-trained version of Fast FullSubNet, a real-time denoising model trained on the Deep Noise Suppression Challenge dataset of 2020 ([DNS-INTERSPEECH-2020](https://github.com/microsoft/DNS-Challenge/tree/interspeech2020/master)).
13
+
14
+ ## How to run
15
+
16
+ https://fullsubnet.readthedocs.io/en/latest/usage/getting_started.html
17
+
18
+ ## Code
19
+
20
+ https://github.com/Audio-WestlakeU/FullSubNet
21
+
22
+ Note: The code doesn't support real-time streaming out of the box. See [issue-67](https://github.com/Audio-WestlakeU/FullSubNet/issues/67) for details.
23
+
24
+ ## Paper
25
+
26
+ [Fast FullSubNet: Accelerate Full-band and Sub-band Fusion Model for Single-channel Speech Enhancement](https://arxiv.org/abs/2212.09019), Xiang Hao, Xiaofei Li
27
+
28
+ > For many speech enhancement applications, a key feature is that system runs on a real-time, latency-sensitive, battery-powered platform, which strictly limits the algorithm latency and computational complexity. In this work, we propose a new architecture named Fast FullSubNet dedicated to accelerating the computation of FullSubNet. Specifically, Fast FullSubNet processes sub-band speech spectra in the mel-frequency domain by using cascaded linear-to-mel full-band, sub-band, and mel-to-linear full-band models such that frequencies involved in the sub-band computation are vastly reduced. After that, a down-sampling operation is proposed for the sub-band input sequence to further reduce the computational complexity along the time axis. Experimental results show that, compared to FullSubNet, Fast FullSubNet has only 13\% computational complexity and 16\% processing time, and achieves comparable or even better performance.
29
+
30
+ ## Performance
31
+
32
+ | | With Reverb |   |   |   | No Reverb |   |   |
33
+ -- | -- | -- | -- | -- | -- | -- | --
34
+ Method | WB-PESQ | NB-PESQ | SI-SDR | STOI | WB-PESQ | NB-PESQ | SI-SDR | STOI
35
+ Fast FullSubNet (118 Epochs) | 2.882 | 3.42 | 15.33 | 0.9233 | 2.694 | 3.222 | 16.34 | 0.9571
36
+ [FullSubNet (58 Epochs)](https://github.com/Audio-WestlakeU/FullSubNet/releases/tag/v0.2) (just for comparison) | 2.987 | 3.496 | 15.756 | 0.926 | 2.889 | 3.385 | 17.635 | 0.964
Fast-FullSubNet/models/Fast-FullSubNet/fast_fullsubnet_best_model_118epochs.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:30bb10f5c801c1e4a046eb210fed4296ae9b757b9ae787f19b219dc16e513359
3
+ size 82213668
Fast-FullSubNet/models/Fast-FullSubNet/source.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ https://huggingface.co/fronx/Fast-FullSubNet
FullSubNet+/FullSubNet+. Channel Attention FullSubNet with Complex Spectrograms for Speech Enhancement.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:246a144c939ce4916c25f3dbfd7d8ae4cf16a69f52c71871406680a8def00e10
3
+ size 335624
FullSubNet+/Learnable spectral dimension compression mapping for full-band speech enhancement.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4efffb0b2dc3888045736a1adb3b36405803eb21dcb75759e6a4e4db0a16ecee
3
+ size 4269048
FullSubNet+/code/FullSubNet-plus [freds0] +2 -3.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:2cee801e6ffb90f9aeee09abe838f5f231b806c551a6312d85089a945a0a96ae
3
+ size 598296
FullSubNet+/code/FullSubNet-plus-optimizations.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:107ab85530457c282a9f10c06591db46f3938d3b8d9a7f2869e124a6dbb786ca
3
+ size 622387
FullSubNet+/code/FullSubNet-plus.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9e5165f9b737bdc5781c334193547bbea45eb17864acccbd15829a801106327a
3
+ size 601150
FullSubNet+/models/best_model.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:cb628b2f3fcf8e078dd720ff04ae9fe14ef9b605ace349e9e8bf1bc1dee15032
3
+ size 104438377
FullSubNet/FullSubNet. A Full-Band and Sub-Band Fusion Model for Real-Time Single-Channel Speech Enhancement.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:1417d6a908f52da1b7a298a4bbc794910e2f0f4cc3315d7452e17205f61b4b30
3
+ size 490412
FullSubNet/Improving the Speech Enhancement Model with Discrete Wavelet Transform Sub-Band Features in Adaptive FullSubNet.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:404e510e532509c2ad84e8bcd90f1d74702d87db13d6c4530de322bbe00b0486
3
+ size 6152557
FullSubNet/Speech Enhancement with Fullband-Subband Cross-Attention Network.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5cb9b47c29932d7e9f2681757f372d1d81583004242d01c9ff5139c04fa11233
3
+ size 1331602
FullSubNet/code/FullSubNet (original)/FullSubNet.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9ffbc7dee23bc8e833e075f165f38153e092004f24d8691f2cf89bbdd0f0ea70
3
+ size 1560708
FullSubNet/code/FullSubNet (original)/v0.2/Checkpoints.txt ADDED
@@ -0,0 +1,23 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ This page has two released model checkpoints. All checkpoints include "model_state_dict", "optimizer_state_dict", and some other meta information.
2
+
3
+ The first model checkpoint is the original model checkpoint at the 58th epoch. The performance is shown in this table:
4
+
5
+ --------------------------------------------------------------------------------------------------------------
6
+ | With Reverb | No Reverb
7
+ --------------------------------------------------------------------------------------------------------------
8
+ Method | WB-PESQ NB-PESQ SI-SDR STOI | WB-PESQ NB-PESQ SI-SDR STOI
9
+ --------------------------------------------------------------------------------------------------------------
10
+ FullSubNet | 2.987 3.496 15.756 0.926 | 2.889 3.385 17.635 0.964
11
+ --------------------------------------------------------------------------------------------------------------
12
+
13
+ In addition, some people are interested in the performance when using cumulative normalization. The below one is a pre-trained FullSubNet using cumulative normalization:
14
+
15
+ -------------------------------------------------------------------------------------------------------------------------
16
+ | With Reverb | No Reverb
17
+ -------------------------------------------------------------------------------------------------------------------------
18
+ Method | WB-PESQ NB-PESQ SI-SDR STOI | WB-PESQ NB-PESQ SI-SDR STOI
19
+ -------------------------------------------------------------------------------------------------------------------------
20
+ FullSubNet (Cumulative Norm) | 2.978 3.503 15.820 0.928 | 2.863 3.376 17.913 0.964
21
+ -------------------------------------------------------------------------------------------------------------------------
22
+
23
+ If you want to inference or fine-tune based on these checkpoints, please check the usage in the documents.
FullSubNet/code/FullSubNet (original)/v0.2/FullSubNet-0.2.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6cd2c1097506f1d9dfd0c8102fde6c8a37c62cdc95e1f2749c97f9ea00ae4cdd
3
+ size 580441
FullSubNet/code/FullSubNet (original)/v0.2/RIR.Multichannel.Impulse.Response.Database.+.The.REVERB.challenge.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:4d7dd5bf6212f334d5dabc90f730f538e1be194403257eff01c847648a157e42
3
+ size 11201457
FullSubNet/code/FullSubNet (original)/v0.2/Room Impulse Responses.txt ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ As mentioned in the paper, the room impulse responses (RIRs) come from the Multichannel Impulse Response Database and the Reverb Challenge dataset. Please download the zip package "RIR (Multichannel Impulse Response Database + The REVERB challenge).zip" if you would like to retrain the FullSubNet.
2
+
3
+ Note that the zip package includes a folder "rir" and a file "rir.txt." The folder "rir" contains all separated single-channel RIRs extracted from the above two datasets. The suffix (e.g., "m_") of the filename is the index of a microphone. The file "rir.txt" is just a path list of all RIRs. Please modify it to fit your case before you use it.
4
+
5
+ For some cases, if you would like to extract channel by yourself, you can download these RIRs from pages:
6
+
7
+ 1. Multichannel Impulse Response Database: https://www.eng.biu.ac.il/~gannot/RIR_DATABASE
8
+ 2. The REVERB challenge data: https://reverb2014.dereverberation.com/tools/reverb_tools_for_Generate_mcTrainData.tgz and https://reverb2014.dereverberation.com/tools/reverb_tools_for_Generate_SimData.tgz
FullSubNet/code/FullSubNet (original)/v0.2/cum_fullsubnet_best_model_218epochs.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d08d09107eb276b8dc3d2d9fff995f4354a51fa3347125f52f8b9aea7c339f81
3
+ size 67667419
FullSubNet/code/FullSubNet (original)/v0.2/fullsubnet_best_model_58epochs.tar ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:482615e9709d2d6c10ee0d24f9844ad2da2dae3c0cf897f014839c365917a2e0
3
+ size 67669069
FullSubNet/code/FullSubNet (original)/wiki.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:fc6e3dbc278d0fe286f46bda832e96634fdd32e0bb9a5a2334d2ae85240498db
3
+ size 24550
FullSubNet/code/FullSubNetWithASR.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7fdddb69f7410d7dc4b52e1bad1e575a6508a5281ee3f59c8d5b2f7563e5f919
3
+ size 1461366
FullSubNet/code/SE-FullSubNet.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:6e1e1191114dccb9b8e8f8fc834eb233bb5d20fcb2404ee0f1d0ea86e3b7d560
3
+ size 135338
FullSubNet/code/fullsubnet_training.ipynb ADDED
The diff for this file is too large to render. See raw diff
 
Mel-FullSubNet/A Mel Spectrogram Enhancement Paradigm Based on CWT in Speech Synthesis.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:505cb4be82a2cee8257330ea956523a7cdc52a9fad3497bd71a6d734f798381d
3
+ size 615739
Mel-FullSubNet/Mel-FullSubNet. Mel-Spectrogram Enhancement for Improving Both Speech Quality and ASR.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5ba2d828cab305842869d7abdf2ddbd5b6969ff3044a060594a20d45f9bd0c64
3
+ size 304011
Spiking-FullSubNet/DPSNN_Spiking_Neural_Network_for_Low-Latency_Strea.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5a7a179e0b86e06c3a6d86bbab16f29b5ae287f36aeb2bfda0797649da7d3371
3
+ size 2961063
Spiking-FullSubNet/Documentation.txt ADDED
@@ -0,0 +1 @@
 
 
1
+ https://haoxiangsnr.github.io/spiking-fullsubnet
Spiking-FullSubNet/Spiking-FullSubNet.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:8cc6a345dcc22d5a7f3643d7195f37a54bbd89dae85ea5e886747d4955167a69
3
+ size 2261042
Spiking-FullSubNet/Towards Ultra-Low-Power Neuromorphic Speech Enhancement with Spiking-FullSubNet.pdf ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:21ed3dd76d221b13783cdee50a1d92d504d64ef094e4126e8b0b60849e1e8372
3
+ size 5157486
Spiking-FullSubNet/code/spiking-fullsubnet-inference.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:14046b6cd280b9173ed7e2cdeb5906653234a5e6f26389c8950ee2d2a6c93ea4
3
+ size 34166984
Spiking-FullSubNet/code/spiking-fullsubnet.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:bd4f1cceb11d28ceff7bd195cb7916d0bac9019cdc499c57bd545068695856f0
3
+ size 333822299
Spiking-FullSubNet/data/spiking-fullsubnet-data.zip ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:959500bd0aa6943131c3902056219e438c9fe3254fc1cc9e8385b167ea166fd6
3
+ size 6480724
Spiking-FullSubNet/data/validation_set.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:41c3c379a354388020a72c3564ed851c450b31467a6795a63906fd13c9d7f39b
3
+ size 735899769