TeslaPoseGen Dataset

TeslaPoseGen dataset provides images of a Tesla Model 3 car from various poses at various locations. Along with the images, we provide paired pixel-level segmentation mask for car in each image. Furthermore, background images from each location is provided to assist the development of pose-conditioned image generation stream of research.

Figure 1: TeslaPoseGen {location: sunnyvale} | left-to-right -> background, car, mask

Figure 2: TeslaPoseGen {location: santa-clara} | left-to-right -> background, car, mask

A few samples from the dataset is shown in figure 1 and 2, and the data distribution for this dataset is given in table 1.

Table 1: Data distribution for TeslaPoseGen dataset

To download the dataset, please follow the instructions below, and if you use this dataset in your research, please cite this work.

pip install gdown

gdown https://drive.google.com/uc?id=1FMrdt5PdWZFlw2qED7OMPvJnqrRd-44k

unzip TeslaPoseGen.zip

citation

@misc{shrivastava2021posegen,

title={PoseGen: Pose-Conditioned Image Generation},

author={Shubham Shrivastava, Amir Ziai},

year={2021},

url={https://github.com/towardsautonomy/PoseGen}

}