XXXXing commited on
Commit
66ba2bc
·
verified ·
1 Parent(s): 4436059

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +12 -20
README.md CHANGED
@@ -24,8 +24,14 @@ size_categories:
24
  >
25
  > IEEE Robotics and Automation Letters (RAL), 2025
26
 
 
 
 
 
27
  **Mimir** is a hierarchical goal-driven diffusion model for end-to-end autonomous driving. It improves upon **GoalFlow** by explicitly modeling goal uncertainty and accelerating goal inference, enabling more robust and efficient planning. With a ResNet-34 backbone, Mimir achieves 89.3 PDMS and 34.6 EPDMS, demonstrating strong performance in both accuracy and efficiency.
28
 
 
 
29
  This repository contains the **model weights and intermediate data files** required for Mimir's evaluation and training. For the full codebase, please visit the [main Mimir repository](https://github.com/ZebinX/Mimir-Uncertainty-Driving).
30
 
31
  ---
@@ -45,29 +51,15 @@ This repository contains the **model weights and intermediate data files** requi
45
  | `frame_mapping_navhard.yaml` | Frame index mapping for Navhard |
46
  | `global_pose_navhard.npy` | Global pose data for Navhard |
47
 
48
- > Each `*_naviunc/` directory contains `navi.npy` (navigation points) and `unc.npy` (uncertainty estimates).
49
-
50
-
51
-
52
  ## 📊 Results
53
 
54
- ### Navtest (NAVSIM v1)
55
-
56
- | Method | NC↑ | DAC↑ | TTC↑ | EP↑ | PDMS↑ |
57
- |:---|:---:|:---:|:---:|:---:|:---:|
58
- | GoalFlow‡ | 98.2 | 96.4 | 93.8 | 82.6 | 87.9 |
59
- | DiffusionDrive | 98.2 | 96.2 | 94.7 | 82.2 | 88.1 |
60
- | **Mimir** | **98.2** | **97.5** | 94.6 | **83.6** | **89.3** |
61
-
62
- ### Navhard (NAVSIM v2)
63
 
64
- | Method | EPDMS↑ |
65
- |:---|:---:|
66
- | GoalFlow† | 28.9 |
67
- | DiffusionDrive† | 29.1 |
68
- | **Mimir** | **34.6** |
69
 
70
- Please refer to the [paper](https://arxiv.org/pdf/2512.07130) for full comparison tables.
71
 
72
  ---
73
 
@@ -90,4 +82,4 @@ If you find Mimir useful, please consider citing our paper:
90
 
91
  ## Contact
92
 
93
- If you have any questions or suggestions, please open an issue or contact: xzebin@bupt.edu.cn
 
24
  >
25
  > IEEE Robotics and Automation Letters (RAL), 2025
26
 
27
+ <div align="center">
28
+ <img src="./assets/teaser.png" width="75%" />
29
+ </div>
30
+
31
  **Mimir** is a hierarchical goal-driven diffusion model for end-to-end autonomous driving. It improves upon **GoalFlow** by explicitly modeling goal uncertainty and accelerating goal inference, enabling more robust and efficient planning. With a ResNet-34 backbone, Mimir achieves 89.3 PDMS and 34.6 EPDMS, demonstrating strong performance in both accuracy and efficiency.
32
 
33
+
34
+
35
  This repository contains the **model weights and intermediate data files** required for Mimir's evaluation and training. For the full codebase, please visit the [main Mimir repository](https://github.com/ZebinX/Mimir-Uncertainty-Driving).
36
 
37
  ---
 
51
  | `frame_mapping_navhard.yaml` | Frame index mapping for Navhard |
52
  | `global_pose_navhard.npy` | Global pose data for Navhard |
53
 
 
 
 
 
54
  ## 📊 Results
55
 
56
+ <div align="center">
57
+ <img src="./assets/navhard.png" width="49%" />
58
+ <img src="./assets/navtest.png" width="49%" />
 
 
 
 
 
 
59
 
60
+ </div>
 
 
 
 
61
 
62
+ Please refer to the [paper](https://arxiv.org/pdf/2512.07130) for more details.
63
 
64
  ---
65
 
 
82
 
83
  ## Contact
84
 
85
+ If you have any questions or suggestions, please open an issue or contact: xzebin@bupt.edu.cn