Lab Group
header-image_0001_layer_4.jpg
header-image_0003_layer_0.jpg
header-image_0002_layer_1.jpg

Two papers accepted to CVPR 2020

Papers on Non-Adversarial Video Synthesis and Multi-source Adaptation using Hypothesis Transfer Learning accepted to CVPR 2020.

Papers on Non-Adversarial Video Synthesis with Learned Priors and Camera On-boarding for Person Re-identification using Hypothesis Transfer Learning are accepted to CVPR 2020.

1. The paper, Non-Adversarial Video Synthesis with Learned Priors, in CVPR 2020 proposes an approach to generate videos from latent noise vectors, without any reference input frames by jointly optimizing the input latent space, the weights of a generator and a recurrent neural network without a discriminator through non-adversarial learning.

Title: Non-Adversarial Video Synthesis with Learned Priors. Abhishek Aich*, Akash Gupta*, Rameswar Panda, Rakib Hyder, Salman Asif, Amit K. Roy-Chowdhury, Computer Vision and Pattern Recognition (CVPR), 2020. (* joint first authors)

concept_diagram_miraj

2. The paper, Camera On-boarding for  Person Re-identification using  Hypothesis Transfer Learning, in CVPR  2020 proposes an approach to swiftly  on-board new camera(s) in an existing  person re-id network (i) without having  access to the source camera data, that the original network was trained on and (ii) relying upon only a small amount of labeled data after adding the new camera(s).

Title: Camera On-boarding for Person Re-identification using Hypothesis Transfer Learning. Sk Miraj Ahmed, Aske R. lejbolle, Ramsewar Panda, Amit K. Roy-Chowdhury, Computer Vision and Pattern Recognition (CVPR), 2020.