Anastasis Stathopoulos

I am a final year PhD student in Computer Science at Rutgers University advised by Dimitris Metaxas. Before that, I graduated with a diploma (BS + MEng) in Electical & Computer Engineering (ECE) from the National Technical University of Athens. I have spent the summers of 2020 and 2021 at Amazon Prime Video working on methods for video understanding applications.

My current research interests lie at the intersection between computer vision and machine learning. Specifically, I'm interested in understanding the world in 3D in order to facilitate reconstruction of humans, animals and objects from everyday images and videos.

  |     |     |  


Selected Publications

Score-Guided Diffusion for 3D Human Recovery
Anastasis Stathopoulos, Ligong Han, Dimitris Metaxas
Computer Vision and Pattern Recognition (CVPR), 2024
[project page] [code] [arxiv] [bibtex]

Solving inverse problems for 3D human pose and shape reconstruction with score guidance in the latent space of a diffusion model.

Learning Articulated Shape with Keypoint Pseudo-labels from Web Images
Anastasis Stathopoulos, Georgios Pavlakos, Ligong Han, Dimitris Metaxas
Computer Vision and Pattern Recognition (CVPR), 2023
[project page] [code & data] [arxiv] [bibtex]

Training models for 3D animal recovery with minimal annotations using large-scale collections of web images.

Exploiting Unlabeled Data with Vision and Language Models for Object Detection
Shiyu Zhao, Zhixing Zhang, Samuel Shulter, Long Zhao, Vijay Kumar, Anastasis Stathopoulos, Manmonahan Chandraker, Dimitris Metaxas
European Conference on Computer Vision (ECCV), 2022
[code & data] [arxiv] [bibtex]

Improving open-vocabulary and semi-supervised object detection with pseudo-labels from VLMs.

Dual Projection Generative Adversarial Networks for Conditional Image Generation
Ligong Han, Martin Renqiang Min, Anastasis Stathopoulos, Yu Tian , Ruijiang Gao, Asim Kadav, Dimitris Metaxas
International Conference on Computer Vision (ICCV), 2021
[code] [arxiv] [bibtex]

Improving class separability and sample quality by balancing data and label matching in cGANs.

Credits to Jon Barron and Georgia Gkioxari