add benchmarks
Browse files
README.md
ADDED
|
@@ -0,0 +1,176 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
**VLN-PE Benchmark**
|
| 2 |
+
<style type="text/css">
|
| 3 |
+
.tg {border-collapse:collapse;border-spacing:0;}
|
| 4 |
+
.tg td{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px;
|
| 5 |
+
overflow:hidden;padding:10px 5px;word-break:normal;}
|
| 6 |
+
.tg th{border-color:black;border-style:solid;border-width:1px;font-family:Arial, sans-serif;font-size:14px;
|
| 7 |
+
font-weight:normal;overflow:hidden;padding:10px 5px;word-break:normal;}
|
| 8 |
+
.tg .tg-c3ow{border-color:inherit;text-align:center;vertical-align:top}
|
| 9 |
+
.tg .tg-0pky{border-color:inherit;text-align:left;vertical-align:top}
|
| 10 |
+
.tg .tg-fymr{border-color:inherit;font-weight:bold;text-align:left;vertical-align:top}
|
| 11 |
+
</style>
|
| 12 |
+
<table class="tg"><thead>
|
| 13 |
+
<tr>
|
| 14 |
+
<th class="tg-c3ow" rowspan="2"><span style="font-weight:bold">Model</span></th>
|
| 15 |
+
<th class="tg-0pky" rowspan="2"><span style="font-weight:bold">Dataset/Benchmark</span></th>
|
| 16 |
+
<th class="tg-c3ow" colspan="7"><span style="font-weight:bold">Val Seen</span></th>
|
| 17 |
+
<th class="tg-c3ow" colspan="7"><span style="font-weight:bold">Val Unseen</span></th>
|
| 18 |
+
<th class="tg-fymr" rowspan="2">Download</th>
|
| 19 |
+
</tr>
|
| 20 |
+
<tr>
|
| 21 |
+
<th class="tg-fymr">TL</th>
|
| 22 |
+
<th class="tg-fymr">NE</th>
|
| 23 |
+
<th class="tg-fymr">FR</th>
|
| 24 |
+
<th class="tg-fymr">StR</th>
|
| 25 |
+
<th class="tg-fymr">OS</th>
|
| 26 |
+
<th class="tg-fymr">SR</th>
|
| 27 |
+
<th class="tg-fymr">SPL</th>
|
| 28 |
+
<th class="tg-fymr">TL</th>
|
| 29 |
+
<th class="tg-fymr">NE</th>
|
| 30 |
+
<th class="tg-fymr">FR</th>
|
| 31 |
+
<th class="tg-fymr">StR</th>
|
| 32 |
+
<th class="tg-fymr">OS</th>
|
| 33 |
+
<th class="tg-fymr">SR</th>
|
| 34 |
+
<th class="tg-fymr">SPL</th>
|
| 35 |
+
</tr></thead>
|
| 36 |
+
<tbody>
|
| 37 |
+
<tr>
|
| 38 |
+
<td class="tg-c3ow" colspan="17">Zero-shot transfer evaluation from VLN-CE</td>
|
| 39 |
+
</tr>
|
| 40 |
+
<tr>
|
| 41 |
+
<td class="tg-0pky">Seq2Seq-Full</td>
|
| 42 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 43 |
+
<td class="tg-0pky">8.29</td>
|
| 44 |
+
<td class="tg-0pky">7.59</td>
|
| 45 |
+
<td class="tg-0pky">19.15</td>
|
| 46 |
+
<td class="tg-0pky">3.04</td>
|
| 47 |
+
<td class="tg-0pky">17.63</td>
|
| 48 |
+
<td class="tg-0pky">13.83</td>
|
| 49 |
+
<td class="tg-0pky">11.17</td>
|
| 50 |
+
<td class="tg-0pky">8.28</td>
|
| 51 |
+
<td class="tg-0pky">6.99</td>
|
| 52 |
+
<td class="tg-0pky">13.88</td>
|
| 53 |
+
<td class="tg-0pky">3.79</td>
|
| 54 |
+
<td class="tg-0pky">21.83</td>
|
| 55 |
+
<td class="tg-0pky">15.00</td>
|
| 56 |
+
<td class="tg-0pky">11.99</td>
|
| 57 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/zero_shot/seq2seq" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 58 |
+
</tr>
|
| 59 |
+
<tr>
|
| 60 |
+
<td class="tg-0pky">CMA-Full</td>
|
| 61 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 62 |
+
<td class="tg-0pky">6.51</td>
|
| 63 |
+
<td class="tg-0pky">7.28</td>
|
| 64 |
+
<td class="tg-0pky">17.93</td>
|
| 65 |
+
<td class="tg-0pky">6.23</td>
|
| 66 |
+
<td class="tg-0pky">18.09</td>
|
| 67 |
+
<td class="tg-0pky">15.50</td>
|
| 68 |
+
<td class="tg-0pky">14.00</td>
|
| 69 |
+
<td class="tg-0pky">6.55</td>
|
| 70 |
+
<td class="tg-0pky">7.02</td>
|
| 71 |
+
<td class="tg-0pky">15.07</td>
|
| 72 |
+
<td class="tg-0pky">4.31</td>
|
| 73 |
+
<td class="tg-0pky">19.82</td>
|
| 74 |
+
<td class="tg-0pky">16.04</td>
|
| 75 |
+
<td class="tg-0pky">14.63</td>
|
| 76 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/zero_shot/cma" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 77 |
+
</tr>
|
| 78 |
+
<tr>
|
| 79 |
+
<td class="tg-c3ow" colspan="17">Train on VLN-PE</td>
|
| 80 |
+
</tr>
|
| 81 |
+
<tr>
|
| 82 |
+
<td class="tg-0pky">Seq2Seq</td>
|
| 83 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 84 |
+
<td class="tg-0pky">9.89</td>
|
| 85 |
+
<td class="tg-0pky">7.73</td>
|
| 86 |
+
<td class="tg-0pky">22.19</td>
|
| 87 |
+
<td class="tg-0pky">3.04</td>
|
| 88 |
+
<td class="tg-0pky">30.55</td>
|
| 89 |
+
<td class="tg-0pky">19.60</td>
|
| 90 |
+
<td class="tg-0pky">15.67</td>
|
| 91 |
+
<td class="tg-0pky">9.51</td>
|
| 92 |
+
<td class="tg-0pky">7.91</td>
|
| 93 |
+
<td class="tg-0pky">19.67</td>
|
| 94 |
+
<td class="tg-0pky">3.71</td>
|
| 95 |
+
<td class="tg-0pky">27.62</td>
|
| 96 |
+
<td class="tg-0pky">15.89</td>
|
| 97 |
+
<td class="tg-0pky">12.58</td>
|
| 98 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/fine_tuned/seq2seq" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 99 |
+
</tr>
|
| 100 |
+
<tr>
|
| 101 |
+
<td class="tg-0pky">CMA</td>
|
| 102 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 103 |
+
<td class="tg-0pky">10.23</td>
|
| 104 |
+
<td class="tg-0pky">7.32</td>
|
| 105 |
+
<td class="tg-0pky">23.40</td>
|
| 106 |
+
<td class="tg-0pky">2.43</td>
|
| 107 |
+
<td class="tg-0pky">31.04</td>
|
| 108 |
+
<td class="tg-0pky">21.12</td>
|
| 109 |
+
<td class="tg-0pky">16.15</td>
|
| 110 |
+
<td class="tg-0pky">10.09</td>
|
| 111 |
+
<td class="tg-0pky">7.43</td>
|
| 112 |
+
<td class="tg-0pky">18.63</td>
|
| 113 |
+
<td class="tg-0pky">3.12</td>
|
| 114 |
+
<td class="tg-0pky">31.33</td>
|
| 115 |
+
<td class="tg-0pky">18.78</td>
|
| 116 |
+
<td class="tg-0pky">14.56</td>
|
| 117 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/fine_tuned/cma" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 118 |
+
</tr>
|
| 119 |
+
<tr>
|
| 120 |
+
<td class="tg-0pky">RDP</td>
|
| 121 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 122 |
+
<td class="tg-0pky">12.00</td>
|
| 123 |
+
<td class="tg-0pky">7.10</td>
|
| 124 |
+
<td class="tg-0pky">22.95</td>
|
| 125 |
+
<td class="tg-0pky">3.95</td>
|
| 126 |
+
<td class="tg-0pky">33.43</td>
|
| 127 |
+
<td class="tg-0pky">23.86</td>
|
| 128 |
+
<td class="tg-0pky">17.35</td>
|
| 129 |
+
<td class="tg-0pky">11.76</td>
|
| 130 |
+
<td class="tg-0pky">6.97</td>
|
| 131 |
+
<td class="tg-0pky">18.75</td>
|
| 132 |
+
<td class="tg-0pky">4.58</td>
|
| 133 |
+
<td class="tg-0pky">34.06</td>
|
| 134 |
+
<td class="tg-0pky">21.98</td>
|
| 135 |
+
<td class="tg-0pky">16.44</td>
|
| 136 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/fine_tuned/rdp" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 137 |
+
</tr>
|
| 138 |
+
<tr>
|
| 139 |
+
<td class="tg-0pky">Seq2Seq+</td>
|
| 140 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 141 |
+
<td class="tg-0pky">9.92</td>
|
| 142 |
+
<td class="tg-0pky">7.54</td>
|
| 143 |
+
<td class="tg-0pky">26.11</td>
|
| 144 |
+
<td class="tg-0pky">5.59</td>
|
| 145 |
+
<td class="tg-0pky">31.93</td>
|
| 146 |
+
<td class="tg-0pky">19.58</td>
|
| 147 |
+
<td class="tg-0pky">15.13</td>
|
| 148 |
+
<td class="tg-0pky">10.21</td>
|
| 149 |
+
<td class="tg-0pky">7.64</td>
|
| 150 |
+
<td class="tg-0pky">21.82</td>
|
| 151 |
+
<td class="tg-0pky">5.12</td>
|
| 152 |
+
<td class="tg-0pky">30.47</td>
|
| 153 |
+
<td class="tg-0pky">18.13</td>
|
| 154 |
+
<td class="tg-0pky">14.06</td>
|
| 155 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/fine_tuned/seq2seq_plus" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 156 |
+
</tr>
|
| 157 |
+
<tr>
|
| 158 |
+
<td class="tg-0pky">CMA+</td>
|
| 159 |
+
<td class="tg-0pky">R2R VLN-PE</td>
|
| 160 |
+
<td class="tg-0pky">9.09</td>
|
| 161 |
+
<td class="tg-0pky">6.68</td>
|
| 162 |
+
<td class="tg-0pky">20.21</td>
|
| 163 |
+
<td class="tg-0pky">2.74</td>
|
| 164 |
+
<td class="tg-0pky">37.99</td>
|
| 165 |
+
<td class="tg-0pky">28.72</td>
|
| 166 |
+
<td class="tg-0pky">24.24</td>
|
| 167 |
+
<td class="tg-0pky">8.61</td>
|
| 168 |
+
<td class="tg-0pky">7.11</td>
|
| 169 |
+
<td class="tg-0pky">18.63</td>
|
| 170 |
+
<td class="tg-0pky">4.83</td>
|
| 171 |
+
<td class="tg-0pky">31.55</td>
|
| 172 |
+
<td class="tg-0pky">23.31</td>
|
| 173 |
+
<td class="tg-0pky">19.66</td>
|
| 174 |
+
<td class="tg-0pky"><a href="https://huggingface.co/InternRobotics/VLN-PE/tree/main/r2r/fine_tuned/cma_plus" target="_blank" rel="noopener noreferrer">model</a></td>
|
| 175 |
+
</tr>
|
| 176 |
+
</tbody></table>
|