Panoptic-aware Image-to-Image Translation

no code implementations • 3 Dec 2021 • Liyun Zhang, Photchara Ratsamee, Bowen Wang, Zhaojie Luo, Yuki Uranishi, Manabu Higashida, Haruo Takemura

The panoptic perception (i. e., foreground instances and background semantics of the image scene) is extracted to achieve alignment between object content codes of the input domain and panoptic-level style codes sampled from the target style space, then refined by a proposed feature masking module for sharping object boundaries.

Image-to-Image Translation Object +3

Transferring Domain-Agnostic Knowledge in Video Question Answering

no code implementations • 26 Oct 2021 • Tianran Wu, Noa Garcia, Mayu Otani, Chenhui Chu, Yuta Nakashima, Haruo Takemura

Video question answering (VideoQA) is designed to answer a given question based on a relevant video clip.

Question Answering Transfer Learning +1

REST: Real-to-Synthetic Transform for Illumination Invariant Camera Localization

no code implementations • 26 Mar 2018 • Sota Shoman, Tomohiro Mashita, Alexander Plopski, Photchara Ratsamee, Yuki Uranishi, Haruo Takemura

Despite continuously improving processing power and rendering algorithms, synthesized images do not perfectly match real images of the same scene, i. e. there exists a gap between real and synthesized images that also affects the accuracy of camera localization.

Camera Localization