added collage and updated readme
Browse files- README.md +60 -43
- collage.png +3 -0
README.md
CHANGED
@@ -13,7 +13,7 @@ git lfs install # Make sure you have git-lfs installed (https://git-lfs.com)
|
|
13 |
git clone https://huggingface.co/datasets/Meehai/dronescapes
|
14 |
```
|
15 |
|
16 |
-
Note: the dataset has about
|
17 |
|
18 |
<details>
|
19 |
<summary> <b> Option 2. Generating the dataset from raw videos and basic labels </b>.</summary>
|
@@ -41,18 +41,20 @@ We use the [video-representations-extractor](https://gitlab.com/meehai/video-rep
|
|
41 |
the rest of the labels using pre-traing networks or algoritms.
|
42 |
|
43 |
```
|
44 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=0 vre raw_data/videos/atanasie_DJI_0652_full/atanasie_DJI_0652_full_540p.mp4 -o raw_data/npz_540p/atanasie_DJI_0652_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
45 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=1 vre raw_data/videos/barsana_DJI_0500_0501_combined_sliced_2700_14700/barsana_DJI_0500_0501_combined_sliced_2700_14700_540p.mp4 -o raw_data/npz_540p/barsana_DJI_0500_0501_combined_sliced_2700_14700/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
46 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=2 vre raw_data/videos/comana_DJI_0881_full/comana_DJI_0881_full_540p.mp4 -o raw_data/npz_540p/comana_DJI_0881_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
47 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=3 vre raw_data/videos/gradistei_DJI_0787_0788_0789_combined_sliced_3510_13110/gradistei_DJI_0787_0788_0789_combined_sliced_3510_13110_540p.mp4 -o raw_data/npz_540p/gradistei_DJI_0787_0788_0789_combined_sliced_3510_13110/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
48 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=4 vre raw_data/videos/herculane_DJI_0021_full/herculane_DJI_0021_full_540p.mp4 -o raw_data/npz_540p/herculane_DJI_0021_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
49 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=5 vre raw_data/videos/jupiter_DJI_0703_0704_0705_combined_sliced_10650_21715/jupiter_DJI_0703_0704_0705_combined_sliced_10650_21715_540p.mp4 -o raw_data/npz_540p/jupiter_DJI_0703_0704_0705_combined_sliced_10650_21715/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
50 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=6 vre raw_data/videos/norway_210821_DJI_0015_full/norway_210821_DJI_0015_full_540p.mp4 -o raw_data/npz_540p/norway_210821_DJI_0015_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
51 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=7 vre raw_data/videos/olanesti_DJI_0416_full/olanesti_DJI_0416_full_540p.mp4 -o raw_data/npz_540p/olanesti_DJI_0416_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
52 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=0 vre raw_data/videos/petrova_DJI_0525_0526_combined_sliced_2850_11850/petrova_DJI_0525_0526_combined_sliced_2850_11850_540p.mp4 -o raw_data/npz_540p/petrova_DJI_0525_0526_combined_sliced_2850_11850/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
53 |
-
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=1 vre raw_data/videos/slanic_DJI_0956_0957_combined_sliced_780_9780/slanic_DJI_0956_0957_combined_sliced_780_9780_540p.mp4 -o raw_data/npz_540p/slanic_DJI_0956_0957_combined_sliced_780_9780/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations rgb "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary"
|
54 |
```
|
55 |
|
|
|
|
|
56 |
### 1.2.4 Convert Mask2Former from Mapillary classes to segprop8 classes
|
57 |
|
58 |
Since we are using pre-trained Mask2Former which has either mapillary or COCO panoptic classes, we need to convert them to dronescapes-compatible (8) classes.
|
@@ -174,7 +176,12 @@ As per the split from the paper:
|
|
174 |
The data is in `data/*` (see the `ls` call above, it should match even if you download from huggingface).
|
175 |
|
176 |
## 2.1 Using the provided viewer
|
177 |
-
|
|
|
|
|
|
|
|
|
|
|
178 |
```
|
179 |
python scripts/dronescapes_viewer.py data/test_set_annotated_only/ # or any of the 8 directories in data/
|
180 |
```
|
@@ -184,48 +191,58 @@ python scripts/dronescapes_viewer.py data/test_set_annotated_only/ # or any of t
|
|
184 |
|
185 |
```
|
186 |
[MultiTaskDataset]
|
187 |
-
- Path: '/
|
188 |
-
-
|
189 |
-
- Representations (8): [NpzRepresentation(depth_dpt), NpzRepresentation(depth_sfm_manual202204), NpzRepresentation(edges_dexined), NpzRepresentation(normals_sfm_manual202204), NpzRepresentation(opticalflow_rife), NpzRepresentation(rgb), NpzRepresentation(semantic_mask2former_swin_mapillary_converted), NpzRepresentation(semantic_segprop8)]
|
190 |
- Length: 116
|
|
|
191 |
== Shapes ==
|
192 |
{'depth_dpt': torch.Size([540, 960]),
|
193 |
'depth_sfm_manual202204': torch.Size([540, 960]),
|
|
|
194 |
'edges_dexined': torch.Size([540, 960]),
|
|
|
195 |
'normals_sfm_manual202204': torch.Size([540, 960, 3]),
|
196 |
'opticalflow_rife': torch.Size([540, 960, 2]),
|
197 |
'rgb': torch.Size([540, 960, 3]),
|
198 |
-
'semantic_mask2former_swin_mapillary_converted': torch.Size([540, 960]),
|
199 |
-
'semantic_segprop8': torch.Size([540, 960])
|
|
|
200 |
== Random loaded item ==
|
201 |
-
|
202 |
-
|
203 |
-
|
204 |
-
'
|
205 |
-
'
|
206 |
-
'normals_sfm_manual202204':
|
207 |
-
'opticalflow_rife': tensor[540, 960, 2]
|
208 |
-
'rgb': tensor[540, 960, 3]
|
209 |
-
'semantic_mask2former_swin_mapillary_converted': tensor[540, 960]
|
210 |
-
'semantic_segprop8': tensor[540, 960]
|
|
|
211 |
== Random loaded batch ==
|
212 |
-
{'depth_dpt':
|
213 |
-
'depth_sfm_manual202204':
|
214 |
-
'
|
215 |
-
'
|
216 |
-
'
|
217 |
-
'
|
218 |
-
'
|
219 |
-
'
|
|
|
|
|
|
|
220 |
== Random loaded batch using torch DataLoader ==
|
221 |
-
{'depth_dpt':
|
222 |
-
'depth_sfm_manual202204':
|
223 |
-
'
|
224 |
-
'
|
225 |
-
'
|
226 |
-
'
|
227 |
-
'
|
228 |
-
'
|
|
|
|
|
|
|
229 |
```
|
230 |
</details>
|
231 |
|
|
|
13 |
git clone https://huggingface.co/datasets/Meehai/dronescapes
|
14 |
```
|
15 |
|
16 |
+
Note: the dataset has about 500GB, so it may take a while to clone it.
|
17 |
|
18 |
<details>
|
19 |
<summary> <b> Option 2. Generating the dataset from raw videos and basic labels </b>.</summary>
|
|
|
41 |
the rest of the labels using pre-traing networks or algoritms.
|
42 |
|
43 |
```
|
44 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=0 vre raw_data/videos/atanasie_DJI_0652_full/atanasie_DJI_0652_full_540p.mp4 -o raw_data/npz_540p/atanasie_DJI_0652_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
45 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=1 vre raw_data/videos/barsana_DJI_0500_0501_combined_sliced_2700_14700/barsana_DJI_0500_0501_combined_sliced_2700_14700_540p.mp4 -o raw_data/npz_540p/barsana_DJI_0500_0501_combined_sliced_2700_14700/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
46 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=2 vre raw_data/videos/comana_DJI_0881_full/comana_DJI_0881_full_540p.mp4 -o raw_data/npz_540p/comana_DJI_0881_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
47 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=3 vre raw_data/videos/gradistei_DJI_0787_0788_0789_combined_sliced_3510_13110/gradistei_DJI_0787_0788_0789_combined_sliced_3510_13110_540p.mp4 -o raw_data/npz_540p/gradistei_DJI_0787_0788_0789_combined_sliced_3510_13110/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
48 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=4 vre raw_data/videos/herculane_DJI_0021_full/herculane_DJI_0021_full_540p.mp4 -o raw_data/npz_540p/herculane_DJI_0021_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
49 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=5 vre raw_data/videos/jupiter_DJI_0703_0704_0705_combined_sliced_10650_21715/jupiter_DJI_0703_0704_0705_combined_sliced_10650_21715_540p.mp4 -o raw_data/npz_540p/jupiter_DJI_0703_0704_0705_combined_sliced_10650_21715/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
50 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=6 vre raw_data/videos/norway_210821_DJI_0015_full/norway_210821_DJI_0015_full_540p.mp4 -o raw_data/npz_540p/norway_210821_DJI_0015_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
51 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=7 vre raw_data/videos/olanesti_DJI_0416_full/olanesti_DJI_0416_full_540p.mp4 -o raw_data/npz_540p/olanesti_DJI_0416_full/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
52 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=0 vre raw_data/videos/petrova_DJI_0525_0526_combined_sliced_2850_11850/petrova_DJI_0525_0526_combined_sliced_2850_11850_540p.mp4 -o raw_data/npz_540p/petrova_DJI_0525_0526_combined_sliced_2850_11850/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
53 |
+
VRE_DEVICE=cuda CUDA_VISIBLE_DEVICES=1 vre raw_data/videos/slanic_DJI_0956_0957_combined_sliced_780_9780/slanic_DJI_0956_0957_combined_sliced_780_9780_540p.mp4 -o raw_data/npz_540p/slanic_DJI_0956_0957_combined_sliced_780_9780/ --cfg_path scripts/cfg.yaml --batch_size 3 --n_threads_data_storer 4 --output_dir_exist_mode overwrite --representations "rgb" "opticalflow_rife" "depth_dpt" "edges_dexined" "semantic_mask2former_swin_mapillary" "softseg_gb"
|
54 |
```
|
55 |
|
56 |
+
Note: `depth_sfm`, `normals_sfm` and `depth_ufo` are not available in VRE. Contact us for more info about them.
|
57 |
+
|
58 |
### 1.2.4 Convert Mask2Former from Mapillary classes to segprop8 classes
|
59 |
|
60 |
Since we are using pre-trained Mask2Former which has either mapillary or COCO panoptic classes, we need to convert them to dronescapes-compatible (8) classes.
|
|
|
176 |
The data is in `data/*` (see the `ls` call above, it should match even if you download from huggingface).
|
177 |
|
178 |
## 2.1 Using the provided viewer
|
179 |
+
|
180 |
+
The simplest way to explore the data is to use the [provided notebook](scripts/dronescapes_viewer.ipynb). Upon running
|
181 |
+
it, you should get a collage with all the default tasks, like this: ![Collage](collage.png)
|
182 |
+
|
183 |
+
For a CLI-only method, you can use the provided reader as well:
|
184 |
+
|
185 |
```
|
186 |
python scripts/dronescapes_viewer.py data/test_set_annotated_only/ # or any of the 8 directories in data/
|
187 |
```
|
|
|
191 |
|
192 |
```
|
193 |
[MultiTaskDataset]
|
194 |
+
- Path: '/export/home/proiecte/aux/mihai_cristian.pirvu/datasets/dronescapes/data/test_set_annotated_only'
|
195 |
+
- Tasks (11): [DepthRepresentation(depth_dpt), DepthRepresentation(depth_sfm_manual202204), DepthRepresentation(depth_ufo), ColorRepresentation(edges_dexined), EdgesRepresentation(edges_gb), NpzRepresentation(normals_sfm_manual202204), OpticalFlowRepresentation(opticalflow_rife), ColorRepresentation(rgb), SemanticRepresentation(semantic_mask2former_swin_mapillary_converted), SemanticRepresentation(semantic_segprop8), ColorRepresentation(softseg_gb)]
|
|
|
196 |
- Length: 116
|
197 |
+
- Handle missing data mode: 'fill_none'
|
198 |
== Shapes ==
|
199 |
{'depth_dpt': torch.Size([540, 960]),
|
200 |
'depth_sfm_manual202204': torch.Size([540, 960]),
|
201 |
+
'depth_ufo': torch.Size([540, 960, 1]),
|
202 |
'edges_dexined': torch.Size([540, 960]),
|
203 |
+
'edges_gb': torch.Size([540, 960, 1]),
|
204 |
'normals_sfm_manual202204': torch.Size([540, 960, 3]),
|
205 |
'opticalflow_rife': torch.Size([540, 960, 2]),
|
206 |
'rgb': torch.Size([540, 960, 3]),
|
207 |
+
'semantic_mask2former_swin_mapillary_converted': torch.Size([540, 960, 8]),
|
208 |
+
'semantic_segprop8': torch.Size([540, 960, 8]),
|
209 |
+
'softseg_gb': torch.Size([540, 960, 3])}
|
210 |
== Random loaded item ==
|
211 |
+
{'depth_dpt': tensor[540, 960] n=518400 (2.0Mb) x∈[0.043, 1.000] μ=0.341 σ=0.418,
|
212 |
+
'depth_sfm_manual202204': None,
|
213 |
+
'depth_ufo': tensor[540, 960, 1] n=518400 (2.0Mb) x∈[0.115, 0.588] μ=0.297 σ=0.138,
|
214 |
+
'edges_dexined': tensor[540, 960] n=518400 (2.0Mb) x∈[0.000, 0.004] μ=0.003 σ=0.001,
|
215 |
+
'edges_gb': tensor[540, 960, 1] n=518400 (2.0Mb) x∈[0., 1.000] μ=0.063 σ=0.100,
|
216 |
+
'normals_sfm_manual202204': None,
|
217 |
+
'opticalflow_rife': tensor[540, 960, 2] n=1036800 (4.0Mb) x∈[-0.004, 0.005] μ=0.000 σ=0.000,
|
218 |
+
'rgb': tensor[540, 960, 3] n=1555200 (5.9Mb) x∈[0., 1.000] μ=0.392 σ=0.238,
|
219 |
+
'semantic_mask2former_swin_mapillary_converted': tensor[540, 960, 8] n=4147200 (16Mb) x∈[0., 1.000] μ=0.125 σ=0.331,
|
220 |
+
'semantic_segprop8': tensor[540, 960, 8] n=4147200 (16Mb) x∈[0., 1.000] μ=0.125 σ=0.331,
|
221 |
+
'softseg_gb': tensor[540, 960, 3] n=1555200 (5.9Mb) x∈[0., 0.004] μ=0.002 σ=0.001}
|
222 |
== Random loaded batch ==
|
223 |
+
{'depth_dpt': tensor[5, 540, 960] n=2592000 (9.9Mb) x∈[0.043, 1.000] μ=0.340 σ=0.417,
|
224 |
+
'depth_sfm_manual202204': tensor[5, 540, 960] n=2592000 (9.9Mb) NaN!,
|
225 |
+
'depth_ufo': tensor[5, 540, 960, 1] n=2592000 (9.9Mb) x∈[0.115, 0.588] μ=0.296 σ=0.137,
|
226 |
+
'edges_dexined': tensor[5, 540, 960] n=2592000 (9.9Mb) x∈[0.000, 0.004] μ=0.003 σ=0.001,
|
227 |
+
'edges_gb': tensor[5, 540, 960, 1] n=2592000 (9.9Mb) x∈[0., 1.000] μ=0.063 σ=0.102,
|
228 |
+
'normals_sfm_manual202204': tensor[5, 540, 960, 3] n=7776000 (30Mb) NaN!,
|
229 |
+
'opticalflow_rife': tensor[5, 540, 960, 2] n=5184000 (20Mb) x∈[-0.004, 0.006] μ=0.000 σ=0.000,
|
230 |
+
'rgb': tensor[5, 540, 960, 3] n=7776000 (30Mb) x∈[0., 1.000] μ=0.393 σ=0.238,
|
231 |
+
'semantic_mask2former_swin_mapillary_converted': tensor[5, 540, 960, 8] n=20736000 (79Mb) x∈[0., 1.000] μ=0.125 σ=0.331,
|
232 |
+
'semantic_segprop8': tensor[5, 540, 960, 8] n=20736000 (79Mb) x∈[0., 1.000] μ=0.125 σ=0.331,
|
233 |
+
'softseg_gb': tensor[5, 540, 960, 3] n=7776000 (30Mb) x∈[0., 0.004] μ=0.002 σ=0.001}
|
234 |
== Random loaded batch using torch DataLoader ==
|
235 |
+
{'depth_dpt': tensor[5, 540, 960] n=2592000 (9.9Mb) x∈[0.025, 1.000] μ=0.216 σ=0.343,
|
236 |
+
'depth_sfm_manual202204': tensor[5, 540, 960] n=2592000 (9.9Mb) x∈[0., 1.000] μ=0.562 σ=0.335 NaN!,
|
237 |
+
'depth_ufo': tensor[5, 540, 960, 1] n=2592000 (9.9Mb) x∈[0.100, 0.580] μ=0.290 σ=0.128,
|
238 |
+
'edges_dexined': tensor[5, 540, 960] n=2592000 (9.9Mb) x∈[0.000, 0.004] μ=0.003 σ=0.001,
|
239 |
+
'edges_gb': tensor[5, 540, 960, 1] n=2592000 (9.9Mb) x∈[0., 1.000] μ=0.079 σ=0.116,
|
240 |
+
'normals_sfm_manual202204': tensor[5, 540, 960, 3] n=7776000 (30Mb) x∈[0.000, 1.000] μ=0.552 σ=0.253 NaN!,
|
241 |
+
'opticalflow_rife': tensor[5, 540, 960, 2] n=5184000 (20Mb) x∈[-0.013, 0.016] μ=0.000 σ=0.004,
|
242 |
+
'rgb': tensor[5, 540, 960, 3] n=7776000 (30Mb) x∈[0., 1.000] μ=0.338 σ=0.237,
|
243 |
+
'semantic_mask2former_swin_mapillary_converted': tensor[5, 540, 960, 8] n=20736000 (79Mb) x∈[0., 1.000] μ=0.125 σ=0.331,
|
244 |
+
'semantic_segprop8': tensor[5, 540, 960, 8] n=20736000 (79Mb) x∈[0., 1.000] μ=0.125 σ=0.331,
|
245 |
+
'softseg_gb': tensor[5, 540, 960, 3] n=7776000 (30Mb) x∈[0., 0.004] μ=0.002 σ=0.001}
|
246 |
```
|
247 |
</details>
|
248 |
|
collage.png
ADDED
![]() |
Git LFS Details
|