PSP-HDRI+: A Synthetic Dataset Generator for Pre-Training of Human-Centric Computer Vision Models
Salehe Erfanian Ebadi, Saurav Dhakad, Sanjay Vishwakarma, Chunpu Wang, You-Cyuan Jhang, Maciek Chociej, Adam Crespi, Alex Thaman, Sujoy Ganguly
Code Available — Be the first to reproduce this paper.
ReproduceCode
- github.com/unity-technologies/peoplesanspeopleOfficialIn papernone★ 322
Abstract
We introduce a new synthetic data generator PSP-HDRI+ that proves to be a superior pre-training alternative to ImageNet and other large-scale synthetic data counterparts. We demonstrate that pre-training with our synthetic data will yield a more general model that performs better than alternatives even when tested on out-of-distribution (OOD) sets. Furthermore, using ablation studies guided by person keypoint estimation metrics with an off-the-shelf model architecture, we show how to manipulate our synthetic data generator to further improve model performance.