Projects
Hyperspectral Band Selection
Objective
Select the most important hyperspectral bands for machine
learning tasks
Methodology
Global Affinity Matrix Reconstruction
Results
We proposed a method that identifies the most informative
hyperspectral bands to enhance machine learning performance.
Duration
2022–2023
Feature Selection Verification
Objective
Verify the effectiveness of feature selection
Methodology
Compare clustering accuracy before and after applying feature
selection
Results
We verified that selecting the top 10% of representative
features led to improved clustering accuracy.
Duration
2022–2023
Partial Shape Matching
Objective
Estimate the internal 3D pose of the joining material
Methodology
Divide-and-conquer strategy for partial shape matching using
measured point cloud data
Results
We propose a method that performs partial shape matching
between flake surfaces and joining material surfaces based on point cloud measurements,
enabling more accurate alignment and fitting.
Duration
2022–2023
Component Reassembly Assistance
Objective
Assist with component reassembly
Methodology
Recording and replaying the disassembly sequence with simple
marker detection
Results
We propose a method to support component reassembly by guiding
users through the recorded disassembly order, enabling faster and more accurate
assembly.
Duration
2022–2023
Wine Cork Inspection
Objective
Inspect wine corks for scratches or damage
Methodology
Multi-scale Faster R-CNN
Results
We propose a machine vision system to automatically detect the
presence, severity, and type of scratches or damage on wine corks.
Duration
2022–2023
Center-Symmetrical Object Detection
Objective
Detect center-symmetrical objects and identify their centers in
arbitrary images
Methodology
Gabor wavelet analysis
Results
Our method accurately detects center-symmetrical objects and
their central points from arbitrary images.
Duration
2021–2022
Rat Senseing
Objective
Analyze rat behavior using the optomotor response
Methodology
Contour line curvature analysis
Results
We present a system that automatically extracts the rat’s head
gaze orientation from each frame.
Duration
2017–2018
Profile Face
Objective
Detect profile faces
Methodology
Flipping scheme
Results
Our method adapts a frontal face detector (red boxes) to also
detect profile faces (blue boxes) through a flipping-based approach.
Duration
2015–2017
Multi-view Face
Objective
Detect multi-view faces
Methodology
Flipping scheme
Results
We present a multi-view face detector that automatically
detects both frontal faces (with faster performance) and profile faces (with slightly
slower performance).
Duration
2013–2015
Saury Head/Tail Determination
Objective
Align all saury in the same head-to-tail orientation within
each packaging box
Methodology
Custom head/tail determination rule
Results
Our method uses basic body extraction and head/tail
determination techniques to achieve ultra-high-speed processing suitable for packaging
pipelines.
Duration
2013–2015
Line Filter Design
Objective
Extract defects from a relatively flat (or flat along
x-axis/y-axis/dioginal directions) surface
Methodology
Custom line filter design
Results
We propose a series of line filters tailored to detect defects
on
surfaces that are flat or exhibit directional flatness along the x-axis, y-axis, or
diagonals.
Duration
2014–2015
