---
pretty_name: Cityscapes VPS
tags:
- image
- datasets
- webdataset
- pandas
- unipercept
license: afl-3.0
task_categories:
- depth-estimation
- image-segmentation
- video-classification
- object-detection
size_categories:
- 10K<n<100K
---

# Cityscapes VPS

This dataset is derived from the videos in the *validation* split of the Cityscapes[^1] dataset.
It aggregates the images and metadata from Cityscapes[^1], Cityscapes-VPS[^2] and Cityscapes-DVPS[^3] into a single structured format. 
This comprehensive derivative was created out of the need for a batteries-included variant of the dataset for academic purposes.
Specifically, joining samples from the individual datasets in their original structure (each is organized differently) involves a significant amount of boilerplate code. 

This dataset is relevant to computer vision research areas such as:

- Segmentation
- Depth estimation
- Autonomous driving
- Video understanding

## Overview
The following variables are included.

1. **Images.** The input data captured by the left camera from Cityscapes[^1], in 8-bit format. Every sequence has 30 frames.
2. **Segmentation labels.** Derived from Cityscapes[^1] and Cityscapes-DVPS[^3], these labels provide detailed semantic segmentation and instance segmentation information for 6 frames of every sequence.
3. **Depth maps.** Improved depth information from Cityscapes-DVPS[^3], offering enhanced quality over the disparity package from Cityscapes[^1], provided for the same samples as the segmentation labels above.
4. **Camera calibrations.** Includes the intrinsic and extrinsic parameters provided by Cityscapes[^1] for each sequence.
5. **Vehicle odometry.** Odometry data for each frame, a subset of those provided in Cityscapes[^1].

Files are grouped by split, sequence and frame.
This leads to the following structure:
```text
data
    train
        000000
            000000.image.png
            000000.panoptic.png
            000000.depth.tiff
            000000.vehicle.json
            000000.timestamp.txt
            000001.image.png
            000001.panoptic.png
            000001.depth.tiff
            000001.vehicle.json
            000001.timestamp.txt
        000000.camera.json
        000001
            ...
        000001.camera.json
        ...
    val
        000000
            ...
        000000.camera.json
        ...
    test
        000000
            ...
        000000.camera.json

```

The `data` directory in this repository only contains the segmentation and depth map annotations.
The remaining data should be downloaded from official sources using the provided preparation script.


## Preparation

1. Clone this dataset repository.
```bash
git clone https://huggingface.co./datasets/khwstolle/csvps && cd csvps
```

2. Install the [Cityscapes developer kit](https://github.com/mcordts/cityscapesScripts) and build dependencies using `pip`.
```bash
python -m pip install -r requirements.txt
```

3. Run the preparation script provided in this repository. 
   Note that this may prompt your [Cityscapes account](https://cityscapes-dataset.com/login/) login credentials.
```bash
make prepare
```

4. To convert the `train`, `val` and `test` directories into a `tar` archive for use with [WebDataset](https://github.com/webdataset/webdataset), run the following command:

```bash
make build 
```

## Usage

See `examples.ipynb` for instructions.

## Citation

If you use this dataset in your research, please cite the original 
[Cityscapes](https://cityscapes-dataset.com), 
[Cityscapes-VPS](https://github.com/mcahny/vps), and
[Cityscapes-DVPS](https://github.com/joe-siyuan-qiao/ViP-DeepLab) datasets.



[^1]: Cordts et al., “The Cityscapes Dataset for Semantic Urban Scene Understanding” (CVPR 2016)

[^2]: Kim et al., "Video Panoptic Segmentation" (CVPR 2020)

[^3]: Qiao et al., "Learning Visual Perception with Depth-aware Video Panoptic Segmentation" (CVPR 2021)