Automation & Knowledge Laboratory is a new laboratory that started in April 2020.
We are developing a robot system that can recognize the real world and learn intelligent behavior through machine learning.
Apr.2023: Our paper "Multi Event Localization by Audio-Visual Fusion with Omnidirectional Camera and Microphone Array" was accepted to CVPRW'23!
Jan.2023: Our paper "Cross-Level Distillation and Feature Denoising for Cross-Domain Few-Shot Classification" was accepted to ICLR'23!
Jan.2023: Our paper "H-SAUR: Hypothesize, Simulate, Act, Update, and Repeat for Understanding Object Articulations from Interactions" was accepted to ICRA'23!
Oct.2022: Our paper "EvIs-Kitchen: Egocentric Human Activities Recognition with Video and Inertial Sensor data" was accepted to MMM'23!
Feb.2022: Two papers were accepted to ICRA'22!
Dec.2021: Our paper "Unsupervised Learning of Image Segmentation Based on Differentiable Feature Clustering" got the IEEE Signal Processing Society (SPS) Japan Student Journal Paper Award. Congrats Wonjik Kim!
Apr.2021: Asako Kanezaki was awarded the Research Encouragement Award of the Japan Society of Mechanical Engineers!
Mar.2021: Our paper "Unsupervised Learning of Image Segmentation Based on Differentiable Feature Clustering" got the Telecom Technology Student Award. Congrats Wonjik Kim!
Dec.2020: Our paper "Incremental Multi-view Object Detection from a Moving Camera" was accepted to ACM Multimedia Asia 2020!
Nov.2020: Our paper "Deep Reactive Planning in Dynamic Environments" was accepted to CoRL'20!
Jul.2020: Our paper of "Unsupervised Image Segmentation" was accepted to IEEE TIP!
Jul.2020: Our paper of "Efficient Exploration in Constrained Environments" was accepted to IEEE IROS'20!
Apr.2020: I started my own laboratory at Tokyo Institute of Technology as an associate professor!
Our paper "Visual Object Search by Learning Spatial Context" was accepted to IEEE RA-L with ICRA'20 presentation!
Our paper of "RotationNet" was accepted to IEEE TPAMI!
Our paper of "Salient object detection on hyperspectral images" was accepted to IEEE ICASSP'19!