maxim-igenbergs commited on
Commit
f3ebdd0
·
verified ·
1 Parent(s): 2ba0483

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +9 -22
README.md CHANGED
@@ -12,19 +12,13 @@ tags:
12
  - vit
13
  - attention
14
  datasets:
15
- - custom
16
  ---
17
-
18
  # ViT End-to-End Driving Model
19
-
20
  Vision Transformer (ViT) adapted for end-to-end autonomous driving, trained on the Udacity self-driving car simulator for the bachelor's thesis: Dual-Axis Testing of Visual Robustness and Topological Generalization in Vision-based End-to-End Driving Models.
21
-
22
  ## Model Description
23
-
24
  This model applies the Vision Transformer architecture to the end-to-end driving task. Instead of using convolutional layers, ViT splits the input image into patches and processes them using self-attention mechanisms, allowing the model to capture global dependencies in the visual input.
25
-
26
  ### Architecture
27
-
28
  ```
29
  Input: RGB Image (224 × 224 × 3)
30
 
@@ -44,36 +38,29 @@ MLP Head
44
 
45
  Output: [steering, throttle]
46
  ```
47
-
48
  ## Checkpoints
49
-
50
  | Map | Checkpoint |
51
  |-----|------------|
52
  | GenRoads | `genroads_20251202-152358/` |
53
  | Jungle | `jungle_20251201-132938/` |
54
-
55
  ### Files per Checkpoint
56
-
57
- - `best_model.ckpt` PyTorch model checkpoint
58
- - `meta.json` Training configuration and hyperparameters
59
- - `history.csv` Training/validation metrics per epoch
60
- - `loss_curve.png` — Visualization of training progress
61
-
62
  ## Citation
63
-
64
  ```bibtex
65
- @thesis{igenbergs2025dualaxis,
66
  title={Dual-Axis Testing of Visual Robustness and Topological Generalization in Vision-based End-to-End Driving Models},
67
  author={Igenbergs, Maxim},
68
  school={Technical University of Munich},
69
- year={2025},
70
  type={Bachelor's Thesis}
71
  }
72
  ```
73
-
74
  ## Related
75
-
76
  - [DAVE-2 Driving Model](https://huggingface.co/maxim-igenbergs/dave2)
77
  - [DAVE-2-GRU Driving Model](https://huggingface.co/maxim-igenbergs/dave2-gru)
78
  - [TCP Driving Model](https://huggingface.co/maxim-igenbergs/tcp-carla-repro)
79
- - [Evaluation Runs Dataset](https://huggingface.co/datasets/maxim-igenbergs/thesis-runs)
 
 
12
  - vit
13
  - attention
14
  datasets:
15
+ - maxim-igenbergs/thesis-data
16
  ---
 
17
  # ViT End-to-End Driving Model
 
18
  Vision Transformer (ViT) adapted for end-to-end autonomous driving, trained on the Udacity self-driving car simulator for the bachelor's thesis: Dual-Axis Testing of Visual Robustness and Topological Generalization in Vision-based End-to-End Driving Models.
 
19
  ## Model Description
 
20
  This model applies the Vision Transformer architecture to the end-to-end driving task. Instead of using convolutional layers, ViT splits the input image into patches and processes them using self-attention mechanisms, allowing the model to capture global dependencies in the visual input.
 
21
  ### Architecture
 
22
  ```
23
  Input: RGB Image (224 × 224 × 3)
24
 
 
38
 
39
  Output: [steering, throttle]
40
  ```
 
41
  ## Checkpoints
 
42
  | Map | Checkpoint |
43
  |-----|------------|
44
  | GenRoads | `genroads_20251202-152358/` |
45
  | Jungle | `jungle_20251201-132938/` |
 
46
  ### Files per Checkpoint
47
+ - `best_model.ckpt`: PyTorch model checkpoint
48
+ - `meta.json`: Training configuration and hyperparameters
49
+ - `history.csv`: Training/validation metrics per epoch
50
+ - `loss_curve.png`: Visualization of training progress
 
 
51
  ## Citation
 
52
  ```bibtex
53
+ @thesis{igenbergs2026dualaxis,
54
  title={Dual-Axis Testing of Visual Robustness and Topological Generalization in Vision-based End-to-End Driving Models},
55
  author={Igenbergs, Maxim},
56
  school={Technical University of Munich},
57
+ year={2026},
58
  type={Bachelor's Thesis}
59
  }
60
  ```
 
61
  ## Related
 
62
  - [DAVE-2 Driving Model](https://huggingface.co/maxim-igenbergs/dave2)
63
  - [DAVE-2-GRU Driving Model](https://huggingface.co/maxim-igenbergs/dave2-gru)
64
  - [TCP Driving Model](https://huggingface.co/maxim-igenbergs/tcp-carla-repro)
65
+ - [Training Data](https://huggingface.co/datasets/maxim-igenbergs/thesis-data)
66
+ - [Evaluation Runs](https://huggingface.co/datasets/maxim-igenbergs/thesis-runs)