IEICE Technical Report

Online edition: ISSN 2432-6380

Smart Info-Media Systems

Workshop Date : 2026-10-08 - 2026-10-09 / Issue Date : 2026-10-01

SIS2026-12
A Fundamental Study on Object Hiding Adversarial Attack against Generalizable Neural Radiance Fields
Reo Arimura, Mashiho Mukaida, Satoshi Ono (Kagoshima Univ)
pp. 1 - 5

SIS2026-13
Construction and Evaluation of a Synthetic Dataset for Smoke-Like Non-Uniform Haze Removal
Shuma Gyokusen, Reia Tanaka (Kagoshima Univ.), Yoshiaki Ueda (Ryukoku Univ.), Satoshi Ono, Mashiho Mukaida (Kagoshima Univ.)
pp. 6 - 11

SIS2026-14
A Note on Object-Level Quality Assessment and Selective Refinement of Low-Quality Objects in 4D Scene Reconstruction
Yuiko Uchida, Ren Togo, Keisuke Maeda, Takahiro Ogawa, Miki Haseyama (Hokkaido Univ.)
pp. 12 - 16

SIS2026-15
A Study on a 2D Code Decoding Method Robust to Geometric Distortion -- Validation of the Effectiveness of Implementing a Deep Learning Model Capable of Estimating Version Numbers --
Kazue Niwa, Yuki Hamada, Shoko Hira, Mashiho Mukaida, Satoshi Ono (Kagoshima Univ.)
pp. 17 - 22

SIS2026-16
Development of a Deep Learning-Based Grading Support System for Pears
Shutaro Konuma, Tatsuya Yamazaki (Niigata Univ.)
pp. 23 - 28

SIS2026-17
FPGA Implementation and Performance Evaluation of a Binarized MLP-Mixer
Shota Hayashi, Ryo Kumagai, Shu Takemoto, Yusuke Nozaki, Masaya Yoshikawa (Meijo Univ.)
pp. 29 - 34

SIS2026-18
Conditions for Content-Preserving Output Manipulation of Vision-Language Models via Adversarial Images
Naoto Katakai, Ryo Kumagai, Shu Takemoto, Yusuke Nozaki, Masaya Yoshikawa (Meijo Univ.)
pp. 35 - 40

SIS2026-19
A Note on Pre-retrieval Failure Prediction Using Internal Representations in Zero-shot Composed Image Retrieval
Kenta Uesugi, Naoki Saito, Keisuke Maeda, Takahiro Ogawa, Miki Haseyama (Hokkaido Univ.)
pp. 41 - 45

SIS2026-20
[Invited Talk] Safety Evaluation of Large Language Models
Tetsuro Takahashi (Kagoshima Univ.)
p. 46

SIS2026-21
[Tutorial Lecture] Bridging Heterogeneous Modalities via Shared Representation Learning
Naoki Saito (Hokkaido Univ.)
pp. 47 - 52

SIS2026-22
Bayesian Regression Analysis of the Relationship between RMS Sound Pressure and Hurst Exponent in Environmental Noise Time Series -- Comparison of the Free and Fixed Models --
Hideo Shibayama (Shibaura Institute of Technology), Yoshiaki Makabe (Kanagawa Institute of Technology), Kenji Muto (Shibaura Institute of Technology), Tomoaki Kimura (Kanagawa Institute of Technology)
pp. 53 - 58

SIS2026-23
(See Japanese page.)
pp. 59 - 62

SIS2026-24
Underwater Acoustic Communication Using Frequency-Domain Narrowband Denoising Autoencoder
Yuki Utashiro, Takanobu Kirihara (Kitami Inst. of Tech.), Shingo Yoshizawa (Chitose Inst. of Sci. and Tech.)
pp. 63 - 67

SIS2026-25
Analysis and Measurement of Acoustic Wave Propagation in Shallow Water for Underwater Acoustic Communication
Ryota Gotou, Takanobu Kirihara (Kitami Inst. of Tech.), Shingo Yoshizawa (Chitose Inst. of Sci. and Tech.)
pp. 68 - 72

SIS2026-26
Noise Reduction in Underwater Acoustic communication Using Convolutional Neural Networks
Reiji Sato, Takanobu Kirihara (Kitami Inst. of Tech.), Shingo Yoshizawa (Chitose Inst. of Sci. and Tech.)
pp. 73 - 78

SIS2026-27
A Proposal of Exposure Fusion Method Based on Brightness, Contrast, and Saturation Adjustments Using S-shape Functions
Seiichi Kojima, Noriaki Suetake (Yamaguchi Univ.)
pp. 79 - 84

SIS2026-28
A Proposal of Chrominance Correction for Protanopia and Deuteranopia in YPQ Color Space
Ayaka Fujita (Yamaguchi Univ.), Takanori Koga (Kindai Univ.), Noriaki Suetake (Yamaguchi Univ.)
pp. 85 - 90

SIS2026-29
A Proposal of Decolorization Method Using Principal Component Analysis for Color Images
Hayate Shigeoka, Ayaka Fujita, Noritake Noriaki (Yamaguchi Univ.)
pp. 91 - 96

SIS2026-30
A Proposal of Memory-Efficient Edge-Preserving Smoothing Filter Based on Block Division
Motoharu Nakano, Ryotaro Honda (Kagoshima Univ.), Yoshiaki Ueda (Ryukoku Univ.), Satoshi Ono, Mashiho Mukaida (Kagoshima Univ.)
pp. 97 - 102

SIS2026-31
A Proposal for Improving Recognition Performance of Low-Light Images Using Trainable Distribution Transformations of White, Black, and Pure Color Components
Yuto Inoue, Yuto Imamura (Kagoshima Univ.), Yoshiaki Ueda (Ryukoku Univ.), Satoshi Ono, Mashiho Mukaida (Kagoshima Univ.)
pp. 103 - 108

Note: Each article is a technical report without peer review, and its polished version will be published elsewhere.

The Institute of Electronics, Information and Communication Engineers (IEICE), Japan