-
Notifications
You must be signed in to change notification settings - Fork 52
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
Merge pull request #20 from ARM-software/add_micronets
Add MicroNets
- Loading branch information
Showing
37 changed files
with
941 additions
and
0 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
72 changes: 72 additions & 0 deletions
72
models/anomaly_detection/micronet_large/tflite_int8/README.md
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,72 @@ | ||
# MicroNet Large INT8 | ||
|
||
## Description | ||
This is a fully quantized version (asymmetrical int8) of the MicroNet Large model developed by Arm, from the MicroNets paper. It is trained on the 'slide rail' task from http://dcase.community/challenge2020/task-unsupervised-detection-of-anomalous-sounds. | ||
|
||
## License | ||
[Apache-2.0](https://spdx.org/licenses/Apache-2.0.html) | ||
|
||
## Related Materials | ||
### Class Labels | ||
The class labels associated with this model can be created by running the script `get_class_labels.sh`. | ||
|
||
## Network Information | ||
| Network Information | Value | | ||
|---------------------|----------------| | ||
| Framework | TensorFlow Lite | | ||
| SHA-1 Hash | 0b7e7776c79fac28c186b2ba00314a05b7faadbf | | ||
| Size (Bytes) | 442000 | | ||
| Provenance | https://arxiv.org/pdf/2010.11267.pdf | | ||
| Paper | https://arxiv.org/pdf/2010.11267.pdf | | ||
|
||
## Performance | ||
| Platform | Optimized | | ||
|----------|:---------:| | ||
| Cortex-A |:heavy_multiplication_x: | | ||
| Cortex-M |:heavy_check_mark: | | ||
| Mali GPU |:heavy_multiplication_x: | | ||
| Ethos U |:heavy_check_mark: | | ||
|
||
### Key | ||
* :heavy_check_mark: - Will run on this platform. | ||
* :heavy_multiplication_x: - Will not run on this platform. | ||
|
||
## Accuracy | ||
Dataset: Dcase 2020 Task 2 Slide Rail | ||
|
||
| Metric | Value | | ||
|--------|-------| | ||
| AUC | 0.968 | | ||
|
||
## Optimizations | ||
| Optimization | Value | | ||
|--------------|---------| | ||
| Quantization | INT8 | | ||
|
||
## Network Inputs | ||
<table> | ||
<tr> | ||
<th width="200">Input Node Name</th> | ||
<th width="100">Shape</th> | ||
<th width="300">Description</th> | ||
</tr> | ||
<tr> | ||
<td>input</td> | ||
<td>(1, 32, 32, 1)</td> | ||
<td>Input is 64 steps of a Log Mel Spectrogram using 64 mels resized to 32x32.</td> | ||
</tr> | ||
</table> | ||
|
||
## Network Outputs | ||
<table> | ||
<tr> | ||
<th width="200">Output Node Name</th> | ||
<th width="100">Shape</th> | ||
<th width="300">Description</th> | ||
</tr> | ||
<tr> | ||
<td>Identity</td> | ||
<td>(1, 8)</td> | ||
<td>Raw logits corresponding to different machine IDs being anomalous</td> | ||
</tr> | ||
</table> |
3 changes: 3 additions & 0 deletions
3
models/anomaly_detection/micronet_large/tflite_int8/ad_large_int8.tflite
Git LFS file not shown
43 changes: 43 additions & 0 deletions
43
models/anomaly_detection/micronet_large/tflite_int8/definition.yaml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,43 @@ | ||
benchmark: | ||
DCASE 2020 Task 2 Slide rail: | ||
AUC: 0.968 | ||
description: This is a fully quantized version (asymmetrical int8) of the MicroNet | ||
Large model developed by Arm, from the MicroNets paper. It is trained on the 'slide | ||
rail' task from http://dcase.community/challenge2020/task-unsupervised-detection-of-anomalous-sounds. | ||
license: | ||
- Apache-2.0 | ||
network: | ||
file_size_bytes: 442000 | ||
filename: ad_large_int8.tflite | ||
framework: TensorFlow Lite | ||
hash: | ||
algorithm: sha1 | ||
value: 0b7e7776c79fac28c186b2ba00314a05b7faadbf | ||
provenance: https://arxiv.org/pdf/2010.11267.pdf | ||
network_parameters: | ||
input_nodes: | ||
- description: Input is 64 steps of a Log Mel Spectrogram using 64 mels resized | ||
to 32x32. | ||
example_input: | ||
path: models/anomaly_detection/micronet_large/tflite_int8/testing_input/input | ||
name: input | ||
shape: | ||
- 1 | ||
- 32 | ||
- 32 | ||
- 1 | ||
output_nodes: | ||
- description: Raw logits corresponding to different machine IDs being anomalous | ||
name: Identity | ||
shape: | ||
- 1 | ||
- 8 | ||
test_output_path: models/anomaly_detection/micronet_large/tflite_int8/testing_output/Identity | ||
operators: | ||
TensorFlow Lite: | ||
- AVERAGE_POOL_2D | ||
- CONV_2D | ||
- DEPTHWISE_CONV_2D | ||
- RELU6 | ||
- RESHAPE | ||
paper: https://arxiv.org/pdf/2010.11267.pdf |
23 changes: 23 additions & 0 deletions
23
models/anomaly_detection/micronet_large/tflite_int8/get_class_labels.sh
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,23 @@ | ||
# Copyright (C) 2021 Arm Limited or its affiliates. All rights reserved. | ||
# | ||
# SPDX-License-Identifier: Apache-2.0 | ||
# | ||
# Licensed under the Apache License, Version 2.0 (the License); you may | ||
# not use this file except in compliance with the License. | ||
# You may obtain a copy of the License at | ||
# | ||
# www.apache.org/licenses/LICENSE-2.0 | ||
# | ||
# Unless required by applicable law or agreed to in writing, software | ||
# distributed under the License is distributed on an AS IS BASIS, WITHOUT | ||
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
# See the License for the specific language governing permissions and | ||
# limitations under the License. | ||
|
||
#!/usr/bin/env bash | ||
|
||
touch ./labelmappings.txt | ||
echo "id0" >> labelmappings.txt | ||
echo "id2" >> labelmappings.txt | ||
echo "id4" >> labelmappings.txt | ||
echo "id6" >> labelmappings.txt |
3 changes: 3 additions & 0 deletions
3
models/anomaly_detection/micronet_large/tflite_int8/testing_input/input/0.npy
Git LFS file not shown
3 changes: 3 additions & 0 deletions
3
models/anomaly_detection/micronet_large/tflite_int8/testing_output/Identity/0.npy
Git LFS file not shown
72 changes: 72 additions & 0 deletions
72
models/anomaly_detection/micronet_medium/tflite_int8/README.md
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,72 @@ | ||
# MicroNet Medium INT8 | ||
|
||
## Description | ||
This is a fully quantized version (asymmetrical int8) of the MicroNet Medium model developed by Arm, from the MicroNets paper. It is trained on the 'slide rail' task from http://dcase.community/challenge2020/task-unsupervised-detection-of-anomalous-sounds. | ||
|
||
## License | ||
[Apache-2.0](https://spdx.org/licenses/Apache-2.0.html) | ||
|
||
## Related Materials | ||
### Class Labels | ||
The class labels associated with this model can be created by running the script `get_class_labels.sh`. | ||
|
||
## Network Information | ||
| Network Information | Value | | ||
|---------------------|----------------| | ||
| Framework | TensorFlow Lite | | ||
| SHA-1 Hash | ed709fccb1d57393cbc88f36da38a4ab70f97b4a | | ||
| Size (Bytes) | 463792 | | ||
| Provenance | https://arxiv.org/pdf/2010.11267.pdf | | ||
| Paper | https://arxiv.org/pdf/2010.11267.pdf | | ||
|
||
## Performance | ||
| Platform | Optimized | | ||
|----------|:---------:| | ||
| Cortex-A |:heavy_multiplication_x: | | ||
| Cortex-M |:heavy_check_mark: | | ||
| Mali GPU |:heavy_multiplication_x: | | ||
| Ethos U |:heavy_check_mark: | | ||
|
||
### Key | ||
* :heavy_check_mark: - Will run on this platform. | ||
* :heavy_multiplication_x: - Will not run on this platform. | ||
|
||
## Accuracy | ||
Dataset: Dcase 2020 Task 2 Slide Rail | ||
|
||
| Metric | Value | | ||
|--------|-------| | ||
| AUC | 0.9632 | | ||
|
||
## Optimizations | ||
| Optimization | Value | | ||
|--------------|---------| | ||
| Quantization | INT8 | | ||
|
||
## Network Inputs | ||
<table> | ||
<tr> | ||
<th width="200">Input Node Name</th> | ||
<th width="100">Shape</th> | ||
<th width="300">Description</th> | ||
</tr> | ||
<tr> | ||
<td>input</td> | ||
<td>(1, 32, 32, 1)</td> | ||
<td>Input is 64 steps of a Log Mel Spectrogram using 64 mels resized to 32x32.</td> | ||
</tr> | ||
</table> | ||
|
||
## Network Outputs | ||
<table> | ||
<tr> | ||
<th width="200">Output Node Name</th> | ||
<th width="100">Shape</th> | ||
<th width="300">Description</th> | ||
</tr> | ||
<tr> | ||
<td>Identity</td> | ||
<td>(1, 8)</td> | ||
<td>Raw logits corresponding to different machine IDs being anomalous</td> | ||
</tr> | ||
</table> |
3 changes: 3 additions & 0 deletions
3
models/anomaly_detection/micronet_medium/tflite_int8/ad_medium_int8.tflite
Git LFS file not shown
43 changes: 43 additions & 0 deletions
43
models/anomaly_detection/micronet_medium/tflite_int8/definition.yaml
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,43 @@ | ||
benchmark: | ||
DCASE 2020 Task 2 Slide rail: | ||
AUC: 0.9632 | ||
description: This is a fully quantized version (asymmetrical int8) of the MicroNet | ||
Medium model developed by Arm, from the MicroNets paper. It is trained on the 'slide | ||
rail' task from http://dcase.community/challenge2020/task-unsupervised-detection-of-anomalous-sounds. | ||
license: | ||
- Apache-2.0 | ||
network: | ||
file_size_bytes: 463792 | ||
filename: ad_medium_int8.tflite | ||
framework: TensorFlow Lite | ||
hash: | ||
algorithm: sha1 | ||
value: ed709fccb1d57393cbc88f36da38a4ab70f97b4a | ||
provenance: https://arxiv.org/pdf/2010.11267.pdf | ||
network_parameters: | ||
input_nodes: | ||
- description: Input is 64 steps of a Log Mel Spectrogram using 64 mels resized | ||
to 32x32. | ||
example_input: | ||
path: models/anomaly_detection/micronet_medium/tflite_int8/testing_input/input | ||
name: input | ||
shape: | ||
- 1 | ||
- 32 | ||
- 32 | ||
- 1 | ||
output_nodes: | ||
- description: Raw logits corresponding to different machine IDs being anomalous | ||
name: Identity | ||
shape: | ||
- 1 | ||
- 8 | ||
test_output_path: models/anomaly_detection/micronet_medium/tflite_int8/testing_output/Identity | ||
operators: | ||
TensorFlow Lite: | ||
- AVERAGE_POOL_2D | ||
- CONV_2D | ||
- DEPTHWISE_CONV_2D | ||
- RELU6 | ||
- RESHAPE | ||
paper: https://arxiv.org/pdf/2010.11267.pdf |
23 changes: 23 additions & 0 deletions
23
models/anomaly_detection/micronet_medium/tflite_int8/get_class_labels.sh
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Original file line number | Diff line number | Diff line change |
---|---|---|
@@ -0,0 +1,23 @@ | ||
# Copyright (C) 2021 Arm Limited or its affiliates. All rights reserved. | ||
# | ||
# SPDX-License-Identifier: Apache-2.0 | ||
# | ||
# Licensed under the Apache License, Version 2.0 (the License); you may | ||
# not use this file except in compliance with the License. | ||
# You may obtain a copy of the License at | ||
# | ||
# www.apache.org/licenses/LICENSE-2.0 | ||
# | ||
# Unless required by applicable law or agreed to in writing, software | ||
# distributed under the License is distributed on an AS IS BASIS, WITHOUT | ||
# WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | ||
# See the License for the specific language governing permissions and | ||
# limitations under the License. | ||
|
||
#!/usr/bin/env bash | ||
|
||
touch ./labelmappings.txt | ||
echo "id0" >> labelmappings.txt | ||
echo "id2" >> labelmappings.txt | ||
echo "id4" >> labelmappings.txt | ||
echo "id6" >> labelmappings.txt |
3 changes: 3 additions & 0 deletions
3
models/anomaly_detection/micronet_medium/tflite_int8/testing_input/input/0.npy
Git LFS file not shown
3 changes: 3 additions & 0 deletions
3
models/anomaly_detection/micronet_medium/tflite_int8/testing_output/Identity/0.npy
Git LFS file not shown
Oops, something went wrong.