https://github.com/j-millet/cv_project3
Instance segmentation of the cityscapes dataset.
https://github.com/j-millet/cv_project3
computer-vision instance-segmentation mask-rcnn pytorch
Last synced: 3 months ago
JSON representation
Instance segmentation of the cityscapes dataset.
- Host: GitHub
- URL: https://github.com/j-millet/cv_project3
- Owner: j-millet
- Created: 2025-01-28T15:56:21.000Z (8 months ago)
- Default Branch: master
- Last Pushed: 2025-01-29T19:00:31.000Z (8 months ago)
- Last Synced: 2025-05-24T01:43:07.907Z (5 months ago)
- Topics: computer-vision, instance-segmentation, mask-rcnn, pytorch
- Language: Python
- Homepage:
- Size: 476 KB
- Stars: 0
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Instance Segmentation on Cityscapes dataset
Instance segmentation on the cityscapes dataset.
Project for computer vision classes at Poznań University of Technology.
## Data
### Cityscapes
The cityscapes dataset can be downloaded by running ```data/scripts/get_files.sh```.
### DVC
The model weights, tensorboard training logs and the custom dataset can be pulled with DVC:
- Download service account login credentials from us.
- Navigate to the ```data``` folder
- Run ```dvc remote modify --local gcp_bucket credentialpath 'PATH/TO/CREDENTIALS'``` to authorize dvc to pull from google cloud.
- Run ```dvc pull``` to download the data.
## Docker environment
Build
```
docker build -t cv3 .
```
Run
```
docker run \
--ipc=host \
--ulimit memlock=-1 \
--ulimit stack=67108864 \
-v ./data:/app/CV_project3/data \
-p 8501:8501 \
-it --rm cv3
```
Run with CUDA
```
docker run \
--gpus all \
--ipc=host \
--ulimit memlock=-1 \
--ulimit stack=67108864 \
-v ./data:/app/CV_project3/data \
-p 8501:8501 \
-it --rm cv3
```
## Goals table:
| **Task** | **Points** | **Done** |
| --- | --- | --- |
| **Problem** |
| Instance Segmentation | 3 | X |
| Additional loss functions to improve prediction quality | 1 | X |
| **Model** |
| Our own model| 2 | X |
| architecture from the internet trained from scratch (mask_rcnn) | 1 | X |
| **Dataset** |
| Your own part of the dataset (>500 photos) | 1 | X |
| **Training** |
| Data augmentation | 1 | X |
| Testing a few optimizers | 1 | X |
| **Additional points** |
| Tensorboard | 1 | X |
| Docker | 1 | X |
| Streamlit | 1 | X |
| DVC | 2 | X |
| **Total** | **15** | **15** |