1 code implementation • 6 Jun 2024 • Jinqi Luo, Tianjiao Ding, Kwan Ho Ryan Chan, Darshan Thaker, Aditya Chattopadhyay, Chris Callison-Burch, René Vidal
Alignment methods are designed to reduce such undesirable output, via techniques such as fine-tuning, prompt engineering, and representation engineering.
1 code implementation • 8 Jun 2023 • Tianzhe Chu, Shengbang Tong, Tianjiao Ding, Xili Dai, Benjamin David Haeffele, René Vidal, Yi Ma
In this paper, we propose a novel image clustering pipeline that leverages the powerful feature representation of large pre-trained models such as CLIP and cluster images effectively and efficiently at scale.
no code implementations • ICCV 2023 • Tianjiao Ding, Shengbang Tong, Kwan Ho Ryan Chan, Xili Dai, Yi Ma, Benjamin D. Haeffele
We consider the problem of simultaneously clustering and learning a linear representation of data lying close to a union of low-dimensional manifolds, a fundamental task in machine learning and computer vision.
no code implementations • CVPR 2022 • Christina Baek, Ziyang Wu, Kwan Ho Ryan Chan, Tianjiao Ding, Yi Ma, Benjamin D. Haeffele
The principle of Maximal Coding Rate Reduction (MCR$^2$) has recently been proposed as a training objective for learning discriminative low-dimensional structures intrinsic to high-dimensional data to allow for more robust training than standard approaches, such as cross-entropy minimization.
no code implementations • 6 Oct 2021 • Yunchen Yang, Xinyue Zhang, Tianjiao Ding, Daniel P. Robinson, Rene Vidal, Manolis C. Tsakiris
In this paper, we revisit the problem of local optimization in RANSAC.
1 code implementation • CVPR 2018 • Kun Huang, Yifan Wang, Zihan Zhou, Tianjiao Ding, Shenghua Gao, Yi Ma
To this end, we have built a very large new dataset of over 5, 000 images with wireframes thoroughly labelled by humans.
no code implementations • CVPR 2020 • Tianjiao Ding, Yunchen Yang, Zhihui Zhu, Daniel P. Robinson, Rene Vidal, Laurent Kneip, Manolis C. Tsakiris
We revisit robust estimation of homographies over point correspondences between two or three views, a fundamental problem in geometric vision.