Sheng-Hsien Hsieh

Sheng-Hsien Hsieh

Auto-generated profile, based on publicly available information

Sheng-Hsien Hsieh

Electrical Engineering, National Yunlin University of Science and Technology University, Yunlin, Taiwan

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
17 November 2023 in IEEE Transactions on Instrumentation and Measurement
Verified
Journal Article
Closed access
MLMing-Yi Lin
Ming-Yi Lin
SHSheng-Hsien Hsieh
Sheng-Hsien Hsieh
Ching-Han CHENChing-Han CHEN
Ching-Han CHEN
CLChun-Hung Lin
Chun-Hung Lin
To address the limitations of typical coil detection systems and enhance the performance of traditional magnetic field imaging (MFI) systems, we propose a MFI system that uses a $4\times4$ array of anisotropic magnetoresistance (AMR) sensors. The core controller of the proposed system was designed using a pipelined parallelization algorithm and implemented on a high-speed field-programmable gate array (FPGA) in a fully hardware-based architecture. The designed MFI system can achieve parallel acquisition of magnetic field components from the 16-channel AMR sensor array, parallel data processing, parallel data transmission, and parallel computations in a single cycle of only 4 ms. Compared with typical detection systems, the proposed solution not only exhibits superior performance but can also provide dynamic visual representations of sensed information.


17 July 2026
Verified
Conference Paper
Open access
CHChen-Pei Hsu
Chen-Pei Hsu
YCYu-Ju Chen
Yu-Ju Chen
CLChe-Chi Li
Che-Chi Li
YHYu-Chieh Huang
Yu-Chieh Huang
SHSheng-Hsien Hsieh
Sheng-Hsien Hsieh
MLMing-Yi Lin
Ming-Yi Lin

4 February 2026
Verified
Conference Paper
Open access
PSPeicyuan Su
Peicyuan Su
HPHo-Feng Peng
Ho-Feng Peng
CWChengen Wu
Chengen Wu
SHShenghsien Hsieh
Shenghsien Hsieh
MLMing-Yi Lin
Ming-Yi Lin
With the growing need for precision agriculture under Agriculture 5.0, reliable early detection of fruit spoilage has become crucial. Traditional visible-light imaging and manual inspection are limited in detecting internal decay and are highly sensitive to illumination changes. This study presents a multispectral deep-learning pipeline (510–920 nm, 16 bands) using MobileNetV2 for guava health classification. Experiments on small (83 images) and large (2000 images) datasets achieved 100% and 99% accuracy, confirming strong model robustness. To interpret decision behavior, Grad-CAM was employed to visualize spectral attention and analyze error patterns. Results show that the model consistently focuses on short-wavelength regions for rotten samples, whereas misclassifications stem from over-sensitivity to local curve fluctuations and high-contrast boundaries. These findings validate the feasibility of applying CNNs to curve-based multispectral data while emphasizing that visual explanations in non-spatial spectral domains require cautious interpretation.

17 July 2026
Verified
Conference Paper
Open access
YWYu-Cheng Wei
Yu-Cheng Wei
CLChi-Jui Li
Chi-Jui Li
CHChen-Pei Hsu
Chen-Pei Hsu
YCYu-Ju Chen
Yu-Ju Chen
SHSheng-Hsien Hsieh
Sheng-Hsien Hsieh
MLMing-Yi Lin
Ming-Yi Lin
Postharvest oranges are susceptible to decay, causing significant economic losses, while manual inspection remains subjective. This study proposes a nondestructive orange decay identification method combining multispectral reflectance curves with a lightweight MobileNetV2-based CNN. ROI reflectance curves (510–920 nm) from 50 orange samples were acquired via a Solomon™ multispectral camera and converted into standardized images for binary classification (healthy vs. rotten). The model achieved 96.6% accuracy, 99.5% healthy-sample Recall, and 99.4% rotten-sample Precision. Main errors involved 12 false-negatives caused by ROI placement near lesion boundaries. Grad-CAM analysis confirmed the model focused on the 510–650 nm region, learning spectrally meaningful features. These results demonstrate the feasibility of this approach for handheld inspection and smart-agriculture grading, though preprocessing inconsistencies between sets introduced a slight input-distribution shift. However, preprocessing inconsistency between the training and testing sets may have introduced a slight input-distribution shift.

11 May 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
SHSheng-Hsien Hsieh
Sheng-Hsien Hsieh
Taiwan imports approximately 97% of its primary energy, while its centralized power grid has experienced repeated large-scale outage events in recent years, exposing the structural vulnerability of single-point-of-failure architectures. This study presents a seven-month empirical evaluation of a self-designed and self-installed residential photovoltaic-battery energy storage system (PV-BESS) deployed on a four-story townhouse in southern Taiwan. The system comprises 3.56 kWp of crystalline silicon modules and a 16.07 kWh lithium iron phosphate (LFP) battery, configured for self-consumption with CT-based anti-backflow control and automatic islanding capability. Monitored from September 2025 through March 2026, the system generated a cumulative 2,276.8 kWh with a daily average of 11.22 kWh. The global-tilted-irradiance-corrected performance ratio averaged 64.2%, consistent with subtropical residential installations. Grid electricity consumption fell by 65.5% compared with the same period before installation, while electricity costs decreased by 89.4%, a disproportionate saving attributable to the progressive tiered tariff structure. The weighted-average self-sufficiency ratio over six months was approximately 58%, and the estimated carbon dioxide reduction was 1,127 kg. Blackout resilience simulations indicate that under a minimum-survival load of 580 W, the battery alone can sustain household operations for approximately 23.6 hours; with favorable solar recharge, the system can exceed the 72-hour preparedness threshold, reaching 7.1 days under optimal conditions. At a self-installation cost of approximately USD 3,800, the simple payback period is estimated at 9.5 years. The implications for resilience-oriented valuation of residential PV-BESS are discussed.