Upload README.md with huggingface_hub

README.md

@@ -61,22 +61,36 @@ dataset = load_dataset("CRUISEResearchGroup/AnyMo-Bench", "AnyMo-Bench-150-US")
 
 ## Baseline Results
 
-The table below reports IMU-based HAR baseline results from the AnyMo paper. Metrics are percentages.
-
-| Setting | Model | Acc@1 | Acc@5 | Macro-F1 |
-|---|---|---:|---:|---:|
-| Fine150 / Unseen Subject | DeepConvLSTM | 35.3 | 63.0 | 17.2 |
-| Fine150 / Unseen Subject | MantisV2 | 38.5 | 65.2 | 22.8 |
-| Fine150 / Unseen Subject | COMODO | 37.8 | 65.2 | 16.0 |
-| Fine150 / Unseen Subject + Cross Device | DeepConvLSTM | 1.9 | 9.5 | 0.4 |
-| Fine150 / Unseen Subject + Cross Device | MantisV2 | 14.4 | 39.7 | 8.6 |
-| Fine150 / Unseen Subject + Cross Device | COMODO | 24.0 | 50.6 | 8.0 |
-| Core50 / Unseen Subject | DeepConvLSTM | 43.2 | 75.4 | 34.5 |
-| Core50 / Unseen Subject | MantisV2 | 45.8 | 76.8 | 41.3 |
-| Core50 / Unseen Subject | COMODO | 46.2 | 78.8 | 37.3 |
-| Core50 / Unseen Subject + Cross Device | DeepConvLSTM | 1.8 | 12.1 | 0.6 |
-| Core50 / Unseen Subject + Cross Device | MantisV2 | 16.6 | 48.7 | 18.4 |
-| Core50 / Unseen Subject + Cross Device | COMODO | 32.6 | 67.8 | 23.3 |
+The table below reports IMU-based HAR baseline results from the AnyMo paper. Metrics are percentages. The best result in each setting/metric is shown in bold, and the second-best result is underlined.
+
+<table>
+  <thead>
+    <tr>
+      <th>Model</th>
+      <th>Acc@1</th>
+      <th>Acc@5</th>
+      <th>Macro-F1</th>
+    </tr>
+  </thead>
+  <tbody>
+    <tr><td colspan="4"><em>Fine150 / Unseen Subject</em></td></tr>
+    <tr><td>DeepConvLSTM</td><td>35.3</td><td>63.0</td><td><u>17.2</u></td></tr>
+    <tr><td>MantisV2</td><td><strong>38.5</strong></td><td><strong>65.2</strong></td><td><strong>22.8</strong></td></tr>
+    <tr><td>COMODO</td><td><u>37.8</u></td><td><u>65.2</u></td><td>16.0</td></tr>
+    <tr><td colspan="4"><em>Fine150 / Unseen Subject + Cross Device</em></td></tr>
+    <tr><td>DeepConvLSTM</td><td>1.9</td><td>9.5</td><td>0.4</td></tr>
+    <tr><td>MantisV2</td><td><u>14.4</u></td><td><u>39.7</u></td><td><strong>8.6</strong></td></tr>
+    <tr><td>COMODO</td><td><strong>24.0</strong></td><td><strong>50.6</strong></td><td><u>8.0</u></td></tr>
+    <tr><td colspan="4"><em>Core50 / Unseen Subject</em></td></tr>
+    <tr><td>DeepConvLSTM</td><td>43.2</td><td>75.4</td><td>34.5</td></tr>
+    <tr><td>MantisV2</td><td><u>45.8</u></td><td><u>76.8</u></td><td><strong>41.3</strong></td></tr>
+    <tr><td>COMODO</td><td><strong>46.2</strong></td><td><strong>78.8</strong></td><td><u>37.3</u></td></tr>
+    <tr><td colspan="4"><em>Core50 / Unseen Subject + Cross Device</em></td></tr>
+    <tr><td>DeepConvLSTM</td><td>1.8</td><td>12.1</td><td>0.6</td></tr>
+    <tr><td>MantisV2</td><td><u>16.6</u></td><td><u>48.7</u></td><td><u>18.4</u></td></tr>
+    <tr><td>COMODO</td><td><strong>32.6</strong></td><td><strong>67.8</strong></td><td><strong>23.3</strong></td></tr>
+  </tbody>
+</table>
 
 ## Schema