5 papers
Pareto-Optimal RTL Code Generation via Multi-Objective Reinforcement Learning with Large Language Models
Koki Takeshita, Kenji Uebayashi, Hikaru Tomonari +6
Recent studies have proposed reinforcement learning with verifiable rewards (RLVR) methods that enable LLMs to generate high-quality RTL circuits from natural-language specificatio…
Bridging the Sim-to-Real Gap in Semiconductor Visual Program Synthesis via Input Binarization
Yusuke Ohtsubo, Kota Dohi, Koichiro Yawata +2
Precise parametric control over circuit geometry is essential for semiconductor inspection, yet obtaining sufficient real training data remains costly. Although generative models s…
LaSTR: Language-Driven Time-Series Segment Retrieval
Kota Dohi, Harsh Purohit, Tomoya Nishida +6
Effectively searching time-series data is essential for system analysis, but existing methods often require expert-designed similarity criteria or rely on global, series-level desc…
A Comparative Study of Meter Detection Methods for Automated Infrastructure Inspection
Yusuke Ohtsubo, Takuto Sato, Hirohiko Sagawa
In order to read meter values from a camera on an autonomous inspection robot with positional errors, it is necessary to detect meter regions from the image. In this study, we deve…
Semi-Supervised Few-Shot Classification with Deep Invertible Hybrid Models
Yusuke Ohtsubo, Tetsu Matsukawa, Einoshin Suzuki
In this paper, we propose a deep invertible hybrid model which integrates discriminative and generative learning at a latent space level for semi-supervised few-shot classification…