Ruichen Ma commited on
Update README.md
Browse files
README.md
CHANGED
|
@@ -37,7 +37,7 @@ This repository contains the **pre-trained weights** for the paper **"I2E: Real-
|
|
| 37 |
|
| 38 |
We provide pre-trained models for **I2E-CIFAR** and **I2E-ImageNet**. You can download the `.pth` files directly from the **Files and versions** tab in this repository.
|
| 39 |
|
| 40 |
-
<table>
|
| 41 |
<tr>
|
| 42 |
<th>Target Dataset</th>
|
| 43 |
<th align="center">Architecture</th>
|
|
@@ -47,75 +47,75 @@ We provide pre-trained models for **I2E-CIFAR** and **I2E-ImageNet**. You can do
|
|
| 47 |
<!-- CIFAR10-DVS -->
|
| 48 |
<tr>
|
| 49 |
<td rowspan="3" align="center" style="vertical-align: middle;"><strong>CIFAR10-DVS</strong><br>(Real)</td>
|
| 50 |
-
<td align="center">MS-ResNet18</td>
|
| 51 |
-
<td align="center">Baseline</td>
|
| 52 |
-
<td align="center">65.6%</td>
|
| 53 |
</tr>
|
| 54 |
<tr>
|
| 55 |
-
<td align="center">MS-ResNet18</td>
|
| 56 |
-
<td align="center">Transfer-I</td>
|
| 57 |
-
<td align="center">83.1%</td>
|
| 58 |
</tr>
|
| 59 |
<tr>
|
| 60 |
-
<td align="center">MS-ResNet18</td>
|
| 61 |
-
<td align="center">Transfer-II (Sim-to-Real)</td>
|
| 62 |
-
<td align="center"><strong>92.5%</strong></td>
|
| 63 |
</tr>
|
| 64 |
<!-- I2E-CIFAR10 -->
|
| 65 |
<tr>
|
| 66 |
<td rowspan="3" align="center" style="vertical-align: middle;"><strong>I2E-CIFAR10</strong></td>
|
| 67 |
-
<td align="center">MS-ResNet18</td>
|
| 68 |
-
<td align="center">Baseline-I</td>
|
| 69 |
-
<td align="center">85.07%</td>
|
| 70 |
</tr>
|
| 71 |
<tr>
|
| 72 |
-
<td align="center">MS-ResNet18</td>
|
| 73 |
-
<td align="center">Baseline-II</td>
|
| 74 |
-
<td align="center">89.23%</td>
|
| 75 |
</tr>
|
| 76 |
<tr>
|
| 77 |
-
<td align="center">MS-ResNet18</td>
|
| 78 |
-
<td align="center">Transfer-I</td>
|
| 79 |
-
<td align="center"><strong>90.86%</strong></td>
|
| 80 |
</tr>
|
| 81 |
<!-- I2E-CIFAR100 -->
|
| 82 |
<tr>
|
| 83 |
<td rowspan="3" align="center" style="vertical-align: middle;"><strong>I2E-CIFAR100</strong></td>
|
| 84 |
-
<td align="center">MS-ResNet18</td>
|
| 85 |
-
<td align="center">Baseline-I</td>
|
| 86 |
-
<td align="center">51.32%</td>
|
| 87 |
</tr>
|
| 88 |
<tr>
|
| 89 |
-
<td align="center">MS-ResNet18</td>
|
| 90 |
-
<td align="center">Baseline-II</td>
|
| 91 |
-
<td align="center">60.68%</td>
|
| 92 |
</tr>
|
| 93 |
<tr>
|
| 94 |
-
<td align="center">MS-ResNet18</td>
|
| 95 |
-
<td align="center">Transfer-I</td>
|
| 96 |
-
<td align="center"><strong>64.53%</strong></td>
|
| 97 |
</tr>
|
| 98 |
<!-- I2E-ImageNet -->
|
| 99 |
<tr>
|
| 100 |
<td rowspan="4" align="center" style="vertical-align: middle;"><strong>I2E-ImageNet</strong></td>
|
| 101 |
-
<td align="center">MS-ResNet18</td>
|
| 102 |
-
<td align="center">Baseline-I</td>
|
| 103 |
-
<td align="center">48.30%</td>
|
| 104 |
</tr>
|
| 105 |
<tr>
|
| 106 |
-
<td align="center">MS-ResNet18</td>
|
| 107 |
-
<td align="center">Baseline-II</td>
|
| 108 |
-
<td align="center">57.97%</td>
|
| 109 |
</tr>
|
| 110 |
<tr>
|
| 111 |
-
<td align="center">MS-ResNet18</td>
|
| 112 |
-
<td align="center">Transfer-I</td>
|
| 113 |
-
<td align="center">59.28%</td>
|
| 114 |
</tr>
|
| 115 |
<tr>
|
| 116 |
-
<td align="center">MS-ResNet34</td>
|
| 117 |
-
<td align="center">Baseline-II</td>
|
| 118 |
-
<td align="center"><strong>60.50%</strong></td>
|
| 119 |
</tr>
|
| 120 |
</table>
|
| 121 |
|
|
@@ -129,18 +129,20 @@ We provide pre-trained models for **I2E-CIFAR** and **I2E-ImageNet**. You can do
|
|
| 129 |
|
| 130 |
Below is the visualization of the I2E conversion process. We illustrate the high-fidelity conversion from static RGB images to dynamic event streams.
|
| 131 |
|
| 132 |
-
<
|
| 133 |
-
<
|
| 134 |
-
|
| 135 |
-
|
| 136 |
-
|
| 137 |
-
</
|
| 138 |
-
<
|
| 139 |
-
<
|
| 140 |
-
|
| 141 |
-
|
| 142 |
-
|
| 143 |
-
</
|
|
|
|
|
|
|
| 144 |
|
| 145 |
## 💻 Usage & Datasets
|
| 146 |
|
|
@@ -162,5 +164,4 @@ If you find this work or the models useful, please cite our AAAI 2026 paper:
|
|
| 162 |
author={Ma, Ruichen and Meng, Liwei and Qiao, Guanchao and Ning, Ning and Liu, Yang and Hu, Shaogang},
|
| 163 |
journal={arXiv preprint arXiv:2511.08065},
|
| 164 |
year={2025}
|
| 165 |
-
}
|
| 166 |
-
```
|
|
|
|
| 37 |
|
| 38 |
We provide pre-trained models for **I2E-CIFAR** and **I2E-ImageNet**. You can download the `.pth` files directly from the **Files and versions** tab in this repository.
|
| 39 |
|
| 40 |
+
<table border="1">
|
| 41 |
<tr>
|
| 42 |
<th>Target Dataset</th>
|
| 43 |
<th align="center">Architecture</th>
|
|
|
|
| 47 |
<!-- CIFAR10-DVS -->
|
| 48 |
<tr>
|
| 49 |
<td rowspan="3" align="center" style="vertical-align: middle;"><strong>CIFAR10-DVS</strong><br>(Real)</td>
|
| 50 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 51 |
+
<td align="center" style="vertical-align: middle;">Baseline</td>
|
| 52 |
+
<td align="center" style="vertical-align: middle;">65.6%</td>
|
| 53 |
</tr>
|
| 54 |
<tr>
|
| 55 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 56 |
+
<td align="center" style="vertical-align: middle;">Transfer-I</td>
|
| 57 |
+
<td align="center" style="vertical-align: middle;">83.1%</td>
|
| 58 |
</tr>
|
| 59 |
<tr>
|
| 60 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 61 |
+
<td align="center" style="vertical-align: middle;">Transfer-II (Sim-to-Real)</td>
|
| 62 |
+
<td align="center" style="vertical-align: middle;"><strong>92.5%</strong></td>
|
| 63 |
</tr>
|
| 64 |
<!-- I2E-CIFAR10 -->
|
| 65 |
<tr>
|
| 66 |
<td rowspan="3" align="center" style="vertical-align: middle;"><strong>I2E-CIFAR10</strong></td>
|
| 67 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 68 |
+
<td align="center" style="vertical-align: middle;">Baseline-I</td>
|
| 69 |
+
<td align="center" style="vertical-align: middle;">85.07%</td>
|
| 70 |
</tr>
|
| 71 |
<tr>
|
| 72 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 73 |
+
<td align="center" style="vertical-align: middle;">Baseline-II</td>
|
| 74 |
+
<td align="center" style="vertical-align: middle;">89.23%</td>
|
| 75 |
</tr>
|
| 76 |
<tr>
|
| 77 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 78 |
+
<td align="center" style="vertical-align: middle;">Transfer-I</td>
|
| 79 |
+
<td align="center" style="vertical-align: middle;"><strong>90.86%</strong></td>
|
| 80 |
</tr>
|
| 81 |
<!-- I2E-CIFAR100 -->
|
| 82 |
<tr>
|
| 83 |
<td rowspan="3" align="center" style="vertical-align: middle;"><strong>I2E-CIFAR100</strong></td>
|
| 84 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 85 |
+
<td align="center" style="vertical-align: middle;">Baseline-I</td>
|
| 86 |
+
<td align="center" style="vertical-align: middle;">51.32%</td>
|
| 87 |
</tr>
|
| 88 |
<tr>
|
| 89 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 90 |
+
<td align="center" style="vertical-align: middle;">Baseline-II</td>
|
| 91 |
+
<td align="center" style="vertical-align: middle;">60.68%</td>
|
| 92 |
</tr>
|
| 93 |
<tr>
|
| 94 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 95 |
+
<td align="center" style="vertical-align: middle;">Transfer-I</td>
|
| 96 |
+
<td align="center" style="vertical-align: middle;"><strong>64.53%</strong></td>
|
| 97 |
</tr>
|
| 98 |
<!-- I2E-ImageNet -->
|
| 99 |
<tr>
|
| 100 |
<td rowspan="4" align="center" style="vertical-align: middle;"><strong>I2E-ImageNet</strong></td>
|
| 101 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 102 |
+
<td align="center" style="vertical-align: middle;">Baseline-I</td>
|
| 103 |
+
<td align="center" style="vertical-align: middle;">48.30%</td>
|
| 104 |
</tr>
|
| 105 |
<tr>
|
| 106 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 107 |
+
<td align="center" style="vertical-align: middle;">Baseline-II</td>
|
| 108 |
+
<td align="center" style="vertical-align: middle;">57.97%</td>
|
| 109 |
</tr>
|
| 110 |
<tr>
|
| 111 |
+
<td align="center" style="vertical-align: middle;">MS-ResNet18</td>
|
| 112 |
+
<td align="center" style="vertical-align: middle;">Transfer-I</td>
|
| 113 |
+
<td align="center" style="vertical-align: middle;">59.28%</td>
|
| 114 |
</tr>
|
| 115 |
<tr>
|
| 116 |
+
<td align="center" style="vertical-align: middle;"><strong>MS-ResNet34</strong></td>
|
| 117 |
+
<td align="center" style="vertical-align: middle;"><strong>Baseline-II</strong></td>
|
| 118 |
+
<td align="center" style="vertical-align: middle;"><strong>60.50%</strong></td>
|
| 119 |
</tr>
|
| 120 |
</table>
|
| 121 |
|
|
|
|
| 129 |
|
| 130 |
Below is the visualization of the I2E conversion process. We illustrate the high-fidelity conversion from static RGB images to dynamic event streams.
|
| 131 |
|
| 132 |
+
<table border="0" style="width: 100%">
|
| 133 |
+
<tr>
|
| 134 |
+
<td width="25%" align="center"><img src="./assets/original_1.jpg" alt="Original 1" style="width:100%"></td>
|
| 135 |
+
<td width="25%" align="center"><img src="./assets/converted_1.gif" alt="Converted 1" style="width:100%"></td>
|
| 136 |
+
<td width="25%" align="center"><img src="./assets/original_2.jpg" alt="Original 2" style="width:100%"></td>
|
| 137 |
+
<td width="25%" align="center"><img src="./assets/converted_2.gif" alt="Converted 2" style="width:100%"></td>
|
| 138 |
+
</tr>
|
| 139 |
+
<tr>
|
| 140 |
+
<td width="25%" align="center"><img src="./assets/original_3.jpg" alt="Original 3" style="width:100%"></td>
|
| 141 |
+
<td width="25%" align="center"><img src="./assets/converted_3.gif" alt="Converted 3" style="width:100%"></td>
|
| 142 |
+
<td width="25%" align="center"><img src="./assets/original_4.jpg" alt="Original 4" style="width:100%"></td>
|
| 143 |
+
<td width="25%" align="center"><img src="./assets/converted_4.gif" alt="Converted 4" style="width:100%"></td>
|
| 144 |
+
</tr>
|
| 145 |
+
</table>
|
| 146 |
|
| 147 |
## 💻 Usage & Datasets
|
| 148 |
|
|
|
|
| 164 |
author={Ma, Ruichen and Meng, Liwei and Qiao, Guanchao and Ning, Ning and Liu, Yang and Hu, Shaogang},
|
| 165 |
journal={arXiv preprint arXiv:2511.08065},
|
| 166 |
year={2025}
|
| 167 |
+
}
|
|
|