IEICE 2017
130 papers
The 2017 edition of IEICE comprised 130 papers focusing on advancements in computer vision, signal processing, and hardware design. Research themes were oriented toward image enhancement and recognition, including multi-focus image fusion for extended depth of field, adaptive contrast enhancement, and low-light recovery through the decoupling of illumination and noise. Biometric recognition also emerged as a key area, with studies investigating synthetic pre-training and privacy-preserving representation learning.
Methodological trends included the exploration of discriminative learning on the unit hypersphere and the development of multi-stage convolutional neural networks for scale transformation. Beyond image processing, the year's work addressed software analysis through generative models and code deltas, as well as hardware-aware designs for neural activation functions. Specific applications ranged from Instagram popularity analysis to the design of Class E power amplifiers using electro-thermal GaN HEMT models. Prolific contributors to this year's proceedings included Toshihiko Yamasaki, Hideo Saito, and Kiyoharu Aizawa.
Top Research Directions
Research directions most represented by IEICE 2017 papers.
Top Authors
Impactful Papers
Papers from IEICE 2017 that shaped research clusters.
Foundational
Hierarchical Reinforcement Learning Based on Return-Weighted Density Estimation
foundational
The Bottleneck of Fixed Skill Cardinalit
Influential
<i>CLCMiner</i>: Detecting Cross-Language Clones without Intermediates
influential
Learning from the Temporal Evolution of
Hierarchical Reinforcement Learning Based on Return-Weighted Density Estimation
influential
Advantage-Weighted Imitation for Policy
Design of class E power amplifiers by using scalable electro-thermal GaN HEMT model
influential
Mitigating Nanoscale Process Variation i
Papers
« prev Page 3/3