IEEE Technology and Society Magazine - September 2023 - 88

Feature
Analysis of Fisherman
Exploitation in Taiwan
Distant Water Fishing
Periyasami Karthikeyan
National Chung Cheng University
Chiayi 62102, Taiwan
Ayush Pratap
Indian Institute of Technology
Ropar 140001, India
William C.-H. Chu
Tunghai University
Taichung 407224, Taiwan
Pao-Ann Hsiung
National Chung Cheng University
Chiayi 62102, Taiwan
 The analysis of fisherman exploitation in
Taiwan's distant water fishing (DWF) industry is a
critical undertaking to address and combat labor
abuses within this sector. DWF involves fishing
activities conducted in waters beyond Taiwan's
territorial limits, and it is crucial to ensure fair and
ethical treatment of the fishermen involved. The
analysis of fisherman exploitation entails a comprehensive
examination of various factors contributing
to labor abuses. This includes investigating working
conditions, working hours, payment practices,
living conditions aboard fishing vessels, and overall
treatment of the fishermen. Evaluating whether
these practices comply with local and international
labor standards, human rights principles, and relevant
regulations is essential. Human action recognition
(HAR) involves the use of computer vision algorithms
to automatically identify and classify actions
being performed by humans in videos or images.
HAR technology can analyze footage from DWF vessels
and identify actions indicating labor exploitation
or abuse. It is essential to address the issue of
exploitation and abuse in Taiwan's DWF industry
[1], [2]. Lewis et al. [16] explored using " AI " and
face recognition as both a speech act and a technological
tool for monitoring workers or intimidating
migrants and protesters. While this concept has
been implemented in various systems, our primary
Digital Object Identifier 10.1109/MTS.2023.3306539
Date of current version: 21 September 2023.
88
0278-0097/23©2023 IEEE
focus is developing a system to detect and address
long work-hour exploitation.
The scarcity of fishermen is a significant concern
within Taiwan's DWF industry. Implementing
HAR in this industry presents various challenges
that must be addressed. One of these challenges
involves obtaining high-quality video footage or
images essential for accurately recognizing and
classifying fishermen's actions. Without clear and
detailed visuals, the effectiveness of HAR algorithms
may be compromised, hindering the accurate
assessment of fishermen's activities. Another
challenge is the development of robust algorithms
capable of accurately classifying a broad spectrum
of actions. The DWF industry encompasses diverse
tasks, including subtle and hard-to-detect activities.
HAR algorithms must be sophisticated and
capable of effectively categorizing these nuanced
actions to provide meaningful insights. Addressing
these challenges requires a concerted effort
to enhance the quality and availability of video
data for analysis. Investing in advanced imaging
technologies and ensuring standardized recording
practices can significantly improve the accuracy
of HAR systems in the DWF industry [3], [4]. The
Taiwan DWF industry generates data using mobile
applications, closed-circuit television (CCTV), and
an automatic identification system. The generated
data are used for finding labor exploitations. The
different models use data generated by the DWF
industry for finding labor exploitations. Figure 1
IEEE Technology and Society Magazine
http://orcid.org/0000-0001-8977-5520 http://orcid.org/0000-0001-7479-3486
IEEE Technology and Society Magazine - September 2023

Table of Contents for the Digital Edition of IEEE Technology and Society Magazine - September 2023
