Naoshi Kaneko

About

Naoshi Kaneko is a leading researcher at the intersection of human-robot interaction and computer vision, with a primary focus on speech-driven gesture generation for virtual agents and robots. His most impactful work, "Analyzing Input and Output Representations for Speech-Driven Gesture Generation" (152 citations), introduces a novel deep-learning framework that significantly advances data-driven methods for automatically generating conversational gestures from speech. This contribution is critical for creating more natural and engaging human-agent interactions. Kaneko has also made notable contributions to autonomous robotics, including monocular vision-based obstacle detection and global localization from single images in known indoor environments. His work on evaluating gesture generation models using convolutional neural networks further demonstrates his commitment to rigorous, data-driven approaches. With a growing citation impact, Kaneko’s research is shaping the future of how robots and virtual agents perceive and respond to humans, bridging the gap between spoken language and physical expression.

Research Focus

Key Achievements

Top Papers

1. 1
   Analyzing Input and Output Representations for Speech-Driven Gesture Generation

   152 citations · 2019
2. 2
   Evaluation of text-to-gesture generation model using convolutional neural network

   11 citations · 2022
3. 3
   Global Localization from a Single Image in Known Indoor Environments

   6 citations · 2018
4. 4
   Fast Obstacle Detection for Monocular Autonomous Mobile Robots

   5 citations · 2017
5. 5
   On the Importance of Representations for Speech-Driven Gesture Generation

   3 citations · 2019

Key Collaborators

Contact & Links