Skip to content
Aoki Media Sensing Lab.FACULTY OF SCIENCE AND TECHNOLOGYAoki Lab.MEDIA SENSING LAB.
ENJP

Site search

Search by research topic, paper title or author.
HOMESelected for 3 Spotlight Presentations at SSII2026!

Selected for 3 Spotlight Presentations at SSII2026!

🎉 All 3 selected as Spotlight Presentations!

At the 32nd Image Sensing Symposium (SSII2026), held June 10 (Wed) to 12 (Fri), 2026 at Pacifico Yokohama Annex Hall, the Aoki Media Sensing Laboratory will present three research papers. All three were chosen from submitted papers for the Spotlight Sessions (oral presentations) and will also be presented in the Interactive Sessions (poster). We look forward to your discussions at the venue.

Dates: June 10 (Wed) to 12 (Fri), 2026 / Venue: Pacifico Yokohama Annex Hall (https://ssii.jp/）

① 3D Gaussian Reconstruction Using Event Camera with Ray Tracing: From Temporal to Spatial Information

  • Presentation NumberSpotlight [SO1-04] / Interactive [IS1-04]
  • AuthorsKai Kayama1Yoshimitsu Aoki1Shintaro Shiba1,2
  • Affiliation1. Keio University, 2. The University of Tokyo
  • Presentation Date and TimeSpotlight: 6/10 (Wed) 14:00–14:45 (Main Hall) / Interactive: 6/10 (Wed) 14:45–16:15 (IS Poster Venue)

This is the Japanese presentation of GPERT (Geometric-Photometric Event-based 3D Gaussian Ray Tracing), selected as a CVPR 2026 Highlight. Leveraging the high temporal resolution of event cameras, we propose an event-based 3D Gaussian Splatting framework that separates geometry (depth) and snapshot luminance at the event level via ray tracing, achieving high reconstruction performance on real data without relying on pretrained models or COLMAP initialization.

Project：https://e3ai.github.io/gpert/ ／ arXiv：https://arxiv.org/abs/2512.18640

② Overlap-aware ViT Harmonizer for Automatic Color Tone Matching of Different Optical Satellite Images

  • Presentation NumberSpotlight [SO2-03] / Interactive [IS2-03]
  • AuthorsPresented by: Sota Sato1Yuto Shibata1Kazuki Ozeki1Ryo Nakamura2Shintaro Sho2Yuhei Aihara2Yoshimitsu Aoki1
  • AffiliationAffiliations: 1. Keio University, 2. Tenchijin
  • Presentation Date and TimeSpotlight: 6/11 (Thu) 15:30–16:15 (Main Hall) / Interactive: 6/11 (Thu) 16:15–17:45 (IS Poster Venue)

We propose a method to automatically harmonize inconsistent color tones between optical satellite images used in combination, addressing the challenge of variance in color tone across images. Our ViT-based harmonizer focuses on overlapping regions between images to automatically adjust the tones of satellite images taken under different conditions. This is a joint research project between Keio University and Tenchijin, specializing in satellite data analysis.

③ Recognition of Product Assembly States Using Latent Representations of 3D Reconstruction Models

  • Presentation NumberSpotlight [SO2-07] / Interactive [IS2-07]
  • AuthorsPresented by: Kohei Torimi1,2，Jyun-Ting Song3，Kris Kitani3Yoshimitsu Aoki1Takuma Yagi2
  • AffiliationAffiliations: 1. Keio University, 2. National Institute of Advanced Industrial Science and Technology, 3. Carnegie Mellon University
  • Presentation Date and TimeSpotlight: 6/11 (Thu) 15:30–16:15 (Main Hall) / Interactive: 6/11 (Thu) 16:15–17:45 (IS Poster Venue)

We propose a method that uses latent representations obtained from 3D reconstruction models of products to recognize whether the products are correctly assembled (assembly state). This is a collaborative research project between Keio University, the National Institute of Advanced Industrial Science and Technology, and Carnegie Mellon University.

For more details on each presentation, please see the SSII2026 Program . We look forward to seeing you at the venue.