Update README.md
Browse files
README.md
CHANGED
@@ -96,7 +96,7 @@ from renumics import spotlight
|
|
96 |
|
97 |
df = dataset.to_pandas()
|
98 |
simple_layout = datasets.load_dataset_builder("renumics/dcase23-task2-enriched", "dev").config.get_layout(config="simple")
|
99 |
-
spotlight.show(df, dtype={'path': spotlight.Audio, "embeddings_ast-finetuned-audioset-10-10-0.4593": spotlight.Embedding}, layout=simple_layout)
|
100 |
```
|
101 |
You can use the UI to interactively configure the view on the data. Depending on the concrete taks (e.g. model comparison, debugging, outlier detection) you might want to leverage different enrichments and metadata.
|
102 |
|
@@ -107,7 +107,7 @@ In this example we focus on the valve class. We specifically look at normal data
|
|
107 |
from renumics import spotlight
|
108 |
|
109 |
extended_layout = datasets.load_dataset_builder("renumics/dcase23-task2-enriched", "dev").config.get_layout(config="extended")
|
110 |
-
spotlight.show(df, dtype={'path': spotlight.Audio, "embeddings_ast-finetuned-audioset-10-10-0.4593": spotlight.Embedding}, layout=extended_layout)
|
111 |
```
|
112 |
|
113 |
![Analyze DCASE23 Task 2 with Spotlight](data/preview_dcase_2.png "Analyze DCASE23 Task 2 with Spotlight")
|
@@ -156,6 +156,9 @@ a ClassLabel for the label and a ClassLabel for the class.
|
|
156 |
'ast-finetuned-audioset-10-10-0.4593-embeddings': [0.8152204155921936,
|
157 |
1.5862374305725098, ...,
|
158 |
1.7154160737991333]
|
|
|
|
|
|
|
159 |
'anomaly_score_dcase2023_task2_baseline_ae': 8.284389
|
160 |
'prediction_dcase2023_task2_baseline_ae': 0
|
161 |
'prediction_correct_dcase2023_task2_baseline_ae': 1
|
@@ -177,7 +180,8 @@ The length of each audio file is 10 seconds.
|
|
177 |
- `domain`: an integer whose value may be either _0_, indicating that the audio sample is from the _source_ domain, _1_, indicating that the audio sample is from the _target_.
|
178 |
- `class`: an integer as class label.
|
179 |
- `label`: an integer whose value may be either _0_, indicating that the audio sample is _normal_, _1_, indicating that the audio sample contains an _anomaly_.
|
180 |
-
- `
|
|
|
181 |
- `anomaly_score_dcase2023_task2_baseline_ae`: a float representation of the anomaly score according to the baseline implementation
|
182 |
- `prediction_dcase2023_task2_baseline_ae`: an integer whose value may be either _0_, indicating that the audio sample is considered _normal_ by the baseline algorithm, _1_, indicating that the audio sample contains an _anomaly_.
|
183 |
- `prediction_correct_dcase2023_task2_baseline_ae`: an integer whose value may be either _0_, indicating that the baseline prediction is wrong or _1_, indicating that prediction is correct.
|
|
|
96 |
|
97 |
df = dataset.to_pandas()
|
98 |
simple_layout = datasets.load_dataset_builder("renumics/dcase23-task2-enriched", "dev").config.get_layout(config="simple")
|
99 |
+
spotlight.show(df, dtype={'path': spotlight.Audio, "embeddings_ast-finetuned-audioset-10-10-0.4593": spotlight.Embedding, "embeddings_dcase2023_task2_baseline_ae": spotlight.Embedding}, layout=simple_layout)
|
100 |
```
|
101 |
You can use the UI to interactively configure the view on the data. Depending on the concrete taks (e.g. model comparison, debugging, outlier detection) you might want to leverage different enrichments and metadata.
|
102 |
|
|
|
107 |
from renumics import spotlight
|
108 |
|
109 |
extended_layout = datasets.load_dataset_builder("renumics/dcase23-task2-enriched", "dev").config.get_layout(config="extended")
|
110 |
+
spotlight.show(df, dtype={'path': spotlight.Audio, "embeddings_ast-finetuned-audioset-10-10-0.4593": spotlight.Embedding, "embeddings_dcase2023_task2_baseline_ae": spotlight.Embedding}, layout=extended_layout)
|
111 |
```
|
112 |
|
113 |
![Analyze DCASE23 Task 2 with Spotlight](data/preview_dcase_2.png "Analyze DCASE23 Task 2 with Spotlight")
|
|
|
156 |
'ast-finetuned-audioset-10-10-0.4593-embeddings': [0.8152204155921936,
|
157 |
1.5862374305725098, ...,
|
158 |
1.7154160737991333]
|
159 |
+
'embeddings_dcase2023_task2_baseline_ae': [12.602639198303223,
|
160 |
+
16.997364044189453, ...,
|
161 |
+
-0.20931333303451538]
|
162 |
'anomaly_score_dcase2023_task2_baseline_ae': 8.284389
|
163 |
'prediction_dcase2023_task2_baseline_ae': 0
|
164 |
'prediction_correct_dcase2023_task2_baseline_ae': 1
|
|
|
180 |
- `domain`: an integer whose value may be either _0_, indicating that the audio sample is from the _source_ domain, _1_, indicating that the audio sample is from the _target_.
|
181 |
- `class`: an integer as class label.
|
182 |
- `label`: an integer whose value may be either _0_, indicating that the audio sample is _normal_, _1_, indicating that the audio sample contains an _anomaly_.
|
183 |
+
- `embeddings_ast-finetuned-audioset-10-10-0.4593`: an `datasets.Sequence(Value("float32"))` representing audio embeddings that are generated with an [Audio Spectrogram Transformer](https://huggingface.co/docs/transformers/model_doc/audio-spectrogram-transformer#transformers.ASTFeatureExtractor).
|
184 |
+
- `embeddings_dcase2023_task2_baseline_ae`: an `datasets.Sequence(Value("float32"))` representing audio embeddings that are generated with the [**DCASE 2023 Challenge Task 2 Baseline Auto Encoder**](https://github.com/nttcslab/dcase2023_task2_baseline_ae). Dimensionality Reduction with PCA is applied separately for each class with a fit on the respecting training set of samples.
|
185 |
- `anomaly_score_dcase2023_task2_baseline_ae`: a float representation of the anomaly score according to the baseline implementation
|
186 |
- `prediction_dcase2023_task2_baseline_ae`: an integer whose value may be either _0_, indicating that the audio sample is considered _normal_ by the baseline algorithm, _1_, indicating that the audio sample contains an _anomaly_.
|
187 |
- `prediction_correct_dcase2023_task2_baseline_ae`: an integer whose value may be either _0_, indicating that the baseline prediction is wrong or _1_, indicating that prediction is correct.
|