# Model Zoo

A number of pre-trained models are available for download to use in your own project. These models are trained on the datasets listed below and are available in Keras and TensorFlow Lite flatbuffer formats.

## [Signal Denoising Task](https://ambiqai.github.io/heartkit/tasks/denoise/)

The following table provides the latest performance and accuracy results for denoising models.

| NAME                | DATASET           | FS    | DURATION | MODEL          | PARAMS | FLOPS   | METRIC      |
| ------------------- | ----------------- | ----- | -------- | -------------- | ------ | ------- | ----------- |
| __DEN-TCN-SM__      | Synthetic, PTB-XL | 100Hz | 2.5s     | TCN            | 3.3K   | 1.0M    | 18.1 SNR    |
| __DEN-TCN-LG__      | Synthetic, PTB-XL | 100Hz | 2.5s     | TCN            | 6.3K   | 1.8M    | 19.5 SNR    |
| __DEN-PPG-TCN-SM__  | Synthetic         | 100Hz | 2.5s     | TCN            | 3.5K   | 1.1M    | 92.1% COS   |

## [Signal Segmentation Task](https://ambiqai.github.io/heartkit/tasks/segmentation/)

The following table provides the latest performance and accuracy results for ECG segmentation models.

| NAME                 | DATASET                  | FS    | DURATION | # CLASSES | MODEL         | PARAMS | FLOPS   | METRIC    |
| -------------------- | ------------------------ | ----- | -------- | --------- | ------------- | ------ | ------- | --------- |
| __SEG-2-TCN-SM__     | LUDB, Synthetic          | 100Hz | 2.5s     | 2         | TCN           | 2K     | 0.42M   | 96.6% F1  |
| __SEG-4-TCN-SM__     | LUDB, Synthetic          | 100Hz | 2.5s     | 4         | TCN           | 7K     | 2.1M    | 86.3% F1  |
| __SEG-4-TCN-LG__     | LUDB, Synthetic          | 100Hz | 2.5s     | 4         | TCN           | 10K    | 3.9M    | 89.4% F1  |
| __SEG-PPG-2-TCN-SM__ | Synthetic                | 100Hz | 2.5s     | 2         | TCN           | 4K     | 1.43M   | 98.6% F1  |

## [Rhythm Classification Task](https://ambiqai.github.io/heartkit/tasks/rhythm/)

The following table provides the latest performance and accuracy results for rhythm classification models.

| NAME             | DATASET                  | FS    | DURATION | # CLASSES | MODEL          | PARAMS | FLOPS   | METRIC   |
| ---------------- | ------------------------ | ----- | -------- | --------- | -------------- | ------ | ------- | -------- |
| __ARR-2-EFF-SM__ | Icentia11K, PTB-XL, LSAD | 100Hz | 5s       | 2         | EfficientNetV2 | 18K    |  1.2M   | 99.5% F1 |
| __ARR-4-EFF-SM__ | LSAD                     | 100Hz | 5s       | 4         | EfficientNetV2 | 27K    |  1.6M   | 95.9% F1 |

## [Beat Classification Task](https://ambiqai.github.io/heartkit/tasks/beat/)

The following table provides the latest performance and accuracy results for beat classification models.

| NAME            | DATASET    | FS    | DURATION | # CLASSES | MODEL          | PARAMS | FLOPS   | METRIC   |
| --------------- | ---------- | ----- | -------- | --------- | -------------- | ------ | ------- | -------- |
| __BC-2-EFF-SM__ | Icentia11k | 100Hz | 5s       | 2         | EfficientNetV2 | 28K    | 1.8M    | 97.7% F1 |
| __BC-3-EFF-SM__ | Icentia11k | 100Hz | 5s       | 3         | EfficientNetV2 | 41K    | 2.1M    | 92.0% F1 |

## Reproducing Results

Each pre-trained model has a corresponding `configuration.json` file that can be used to reproduce the model and results.

To reproduce a pre-trained rhythm model with configuration file `configuration.json`, run the following command:

```bash title="Terminal"
heartkit -m train -t rhythm -c configuration.json
```

To evaluate the trained rhythm model with configuration file `configuration.json`, run the following command:

```bash title="Terminal"
heartkit -m evaluate -t rhythm -c configuration.json
```