Hugging Face's logo

iky1e
/
DPDFNet4-CoreML

Core ML
Safetensors
audio
speech-enhancement
noise-suppression
streaming
dpdfnet
Model card Files
xet
 Community
DPDFNet4-CoreML
File size: 1,059 Bytes
6866fdc
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
{
  "model_version": "DPDFNet4",
  "dpdfnet_profile": "16khz",
  "sample_rate": 16000,
  "fft_size": 320,
  "hop_size": 160,
  "min_nb_erb_freqs": 1,
  "nb_erb": 32,
  "nb_df": 96,
  "dprnn_num_blocks": 4,
  "dprnn_erb_bins": 8,
  "dprnn_df_bins": 48,
  "encoder_erb_strides": [
    1,
    2,
    2,
    1
  ],
  "decoder_erb_strides": [
    1,
    2,
    2
  ],
  "feature_type": "erb",
  "mask_type": "erb",
  "df_order": 5,
  "df_lookahead": 2,
  "conv_lookahead": 2,
  "conv_ch": 64,
  "emb_hidden_dim": 256,
  "emb_num_layers": 2,
  "df_hidden_dim": 256,
  "df_num_layers": 2,
  "gru_groups": 1,
  "linear_groups": 16,
  "enc_linear_groups": 32,
  "group_shuffle": false,
  "enc_concat": true,
  "df_gru_skip": "groupedlinear",
  "conv_kernel": [
    1,
    3
  ],
  "convt_kernel": [
    1,
    3
  ],
  "conv_kernel_inp": [
    3,
    3
  ],
  "df_pathway_kernel_size_t": 5,
  "lsnr_max": 35,
  "lsnr_min": -15,
  "decoder_upsample_type": "subpixel",
  "source_checkpoint": "dpdfnet4.pth",
  "source_repository": "https://github.com/ceva-ip/DPDFNet"
}