update README
Browse files
README.md
CHANGED
@@ -109,14 +109,12 @@ chmod +x aea.sh
|
|
109 |
```
|
110 |
|
111 |
This script will:
|
112 |
-
- Download 52 AEA video sequences (`.vrs` format) with resume capability
|
113 |
-
- Download corresponding MPS SLAM data (trajectories and point clouds) for camera pose estimation
|
114 |
- Extract RGB images and camera poses from VRS files using `egolifter/scripts/process_project_aria_3dgs.py`
|
115 |
- Remove fisheye distortion and rectify images using `egolifter/scripts/rectify_aria.py`
|
116 |
-
-
|
117 |
-
-
|
118 |
-
-
|
119 |
-
- Save all processed data in `aea/aea_processed/` with structured organization
|
120 |
|
121 |
### Step 3: Verify Processing
|
122 |
After completion, you should have:
|
@@ -127,7 +125,7 @@ After completion, you should have:
|
|
127 |
aea/aea_processed/
|
128 |
βββ loc1_script2_seq1_rec1/
|
129 |
β βββ audio/
|
130 |
-
β β βββ loc1_script2_seq1_rec1.wav #
|
131 |
β βββ video/
|
132 |
β β βββ loc1_script2_seq1_rec1.mp4 # From undistorted frames
|
133 |
β βββ images/ # Undistorted frames
|
|
|
109 |
```
|
110 |
|
111 |
This script will:
|
112 |
+
- Download 52 AEA video sequences (`.vrs` format) and SLAM data with resume capability
|
|
|
113 |
- Extract RGB images and camera poses from VRS files using `egolifter/scripts/process_project_aria_3dgs.py`
|
114 |
- Remove fisheye distortion and rectify images using `egolifter/scripts/rectify_aria.py`
|
115 |
+
- Extract audio from VRS files
|
116 |
+
- Convert undistorted frames to MP4 videos
|
117 |
+
- Save all processed data in `aea/aea_processed/`
|
|
|
118 |
|
119 |
### Step 3: Verify Processing
|
120 |
After completion, you should have:
|
|
|
125 |
aea/aea_processed/
|
126 |
βββ loc1_script2_seq1_rec1/
|
127 |
β βββ audio/
|
128 |
+
β β βββ loc1_script2_seq1_rec1.wav # Extract Audio from VRS
|
129 |
β βββ video/
|
130 |
β β βββ loc1_script2_seq1_rec1.mp4 # From undistorted frames
|
131 |
β βββ images/ # Undistorted frames
|