Name
Abstract | ||
|---|---|---|
•The proposed method prunes unimportant connections in a deep convolutional neural network.•The proposed method identifies the redundant feature maps which produce similar embeddings across classes.•The filters associated with redundant feature maps are being eliminated completely.•The proposed pruning approach does not consider any reconstruction error. |
Year | DOI | Venue |
|---|---|---|
2020 | 10.1016/j.patrec.2020.02.004 | Pattern Recognition Letters |
Keywords | DocType | Volume |
Pruning,SoundNet,Embedding,Response matrix,Acoustic scene classification | Journal | 131 |
ISSN | Citations | PageRank |
0167-8655 | 0 | 0.34 |
References | Authors | |
0 | 3 | |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Arshdeep Singh | 1 | 0 | 2.03 |
| Padmanabhan Rajan | 2 | 22 | 7.63 |
| Arnav Bhavsar | 3 | 69 | 22.05 |