Dong-Hwan Jang — accepted AI-conference papers

2026

PyraTok: Language-Aligned Pyramidal Tokenizer for Video Understanding and Generation

Onkar Susladkar, Tushar Prakash, Adheesh Juvekar, Kiet A. Nguyen, Dong-Hwan Jang, Inderjit S Dhillon +1

CVPR 2026

Discrete video VAEs underpin modern text-to-video generation and video understanding systems, yet existing tokenizers typically learn visual codebooks at a single scale with limited vocabularies and shallow language supervision, leading to poor cross-modal alignment and zero-shot transfer. We introd

Cited by 0 Source ↗code ↗Scholar ↗

2026

RewardFlow: Generate Images by Optimizing What You Reward

Onkar Susladkar, Dong-Hwan Jang, Tushar Prakash, Adheesh Juvekar, Vedant Shah, Ayush Barik +4

CVPR 2026

RewardFlow is a zero-shot, training-free framework for text-guided image editing and generation based on reward-guided Langevin dynamics. We steer pretrained diffusion and flow-matching models at inference time using a diverse set of differentiable rewards, and control their influence with a prompt-

Cited by 0 Source ↗Scholar ↗

2024

Model Stock: All we need is just a few fine-tuned models

Dong-Hwan Jang, Sangdoo Yun, Dongyoon Han*

ECCV 2024oral

"This paper introduces an efficient fine-tuning method for large pre-trained models, offering strong in-distribution (ID) and out-of-distribution (OOD) performance. Breaking away from traditional practices that need a multitude of fine-tuned models for averaging, our approach employs significantly f…

Cited by 29 Source ↗PDF ↗code ↗Scholar ↗

2022

Pooling Revisited: Your Receptive Field Is Suboptimal

Dong-Hwan Jang, Sanghyeok Chu, Joonhyuk Kim, Bohyung Han

CVPR 2022poster

The size and shape of the receptive field determine how the network aggregates local features, and affect the overall performance of a model considerably. Many components in a neural network, such as depth, kernel sizes, and strides for convolution and pooling, influence the receptive field. However…

Cited by 17 PDF ↗Scholar ↗