Shih-Chieh Chang
YOU?
Author Swipe
View article: Therapeutic effects of powered exoskeletal robot-assisted gait training in inpatients in the early stage after stroke: a pilot case-controlled study
Therapeutic effects of powered exoskeletal robot-assisted gait training in inpatients in the early stage after stroke: a pilot case-controlled study Open
This trial was approved by the Institutional Review Board of Chang Gung Medical Foundation (No. 202200822B0).
View article: Synergic Effect of Robot-Assisted Rehabilitation and Antispasticity Therapy: A Narrative Review
Synergic Effect of Robot-Assisted Rehabilitation and Antispasticity Therapy: A Narrative Review Open
Background: Stroke and spinal cord injury are neurological disorders that cause disability and exert tremendous social and economic effects. Robot-assisted training (RAT), which may reduce spasticity, is widely applied in neurorehabilitati…
View article: The Comparison and Interpretation of Machine-Learning Models in Post-Stroke Functional Outcome Prediction
The Comparison and Interpretation of Machine-Learning Models in Post-Stroke Functional Outcome Prediction Open
Prediction of post-stroke functional outcomes is crucial for allocating medical resources. In this study, a total of 577 patients were enrolled in the Post-Acute Care-Cerebrovascular Disease (PAC-CVD) program, and 77 predictors were collec…
View article: Circulating mRNA Expression of astrocyte-Elevated Gene-1 Associated with Treatment Response and Survival in Non-Small Cell Lung Cancer Patients Treated with Pemetrexed
Circulating mRNA Expression of astrocyte-Elevated Gene-1 Associated with Treatment Response and Survival in Non-Small Cell Lung Cancer Patients Treated with Pemetrexed Open
Backgrounds: Astrocyte-elevated gene-1 (AEG-1) functions as an oncogene and regulates angiogenesis in non-small cell lung cancer (NSCLC). In this prospective study, we assessed the values of plasma AEG-1 mRNA expression by liquid biopsy as…
View article: Remix: Rebalanced Mixup
Remix: Rebalanced Mixup Open
Deep image classifiers often perform poorly when training data are heavily class-imbalanced. In this work, we propose a new regularization technique, Remix, that relaxes Mixup's formulation and enables the mixing factors of features and la…
View article: Calibrated BatchNorm: Improving Robustness Against Noisy Weights in Neural Networks.
Calibrated BatchNorm: Improving Robustness Against Noisy Weights in Neural Networks. Open
Analog computing hardware has gradually received more attention by the researchers for accelerating the neural network computations in recent years. However, the analog accelerators often suffer from the undesirable intrinsic noise caused …
View article: Robust Processing-In-Memory Neural Networks via Noise-Aware Normalization
Robust Processing-In-Memory Neural Networks via Noise-Aware Normalization Open
Analog computing hardwares, such as Processing-in-memory (PIM) accelerators, have gradually received more attention for accelerating the neural network computations. However, PIM accelerators often suffer from intrinsic noise in the physic…
View article: Relationship between maximal incremental and high-intensity interval exercise performance in elite athletes
Relationship between maximal incremental and high-intensity interval exercise performance in elite athletes Open
This descriptive study aimed to explore the physiological factors that determine tolerance to exertion during high-intensity interval effort. Forty-seven young women (15-28 years old) were enrolled: 23 athletes from Taiwan national or nati…
View article: AirConcierge: Generating Task-Oriented Dialogue via Efficient Large-Scale Knowledge Retrieval
AirConcierge: Generating Task-Oriented Dialogue via Efficient Large-Scale Knowledge Retrieval Open
Despite recent success in neural task-oriented dialogue systems, developing such a real-world system involves accessing large-scale knowledge bases (KBs), which cannot be simply encoded by neural approaches, such as memory network mechanis…
View article: Relationship between maximal incremental and high-intensity interval exercise performance in elite athletes
Relationship between maximal incremental and high-intensity interval exercise performance in elite athletes Open
It remains unclear whether the number of total bouts to limitation (B lim ) in high-intensity interval testing (HIIT) differs among individuals, no matter if performed at the same relative intensity. This study aimed to explore the physiol…
View article: Learning with Hierarchical Complement Objective
Learning with Hierarchical Complement Objective Open
Label hierarchies widely exist in many vision-related problems, ranging from explicit label hierarchies existed in image classification to latent label hierarchies existed in semantic segmentation. Nevertheless, state-of-the-art methods of…
View article: Improving Adversarial Robustness via Guided Complement Entropy
Improving Adversarial Robustness via Guided Complement Entropy Open
Adversarial robustness has emerged as an important topic in deep learning as carefully crafted attack samples can significantly disturb the performance of a model. Many recent methods have proposed to improve adversarial robustness by util…
View article: Complement Objective Training
Complement Objective Training Open
Learning with a primary objective, such as softmax cross entropy for classification and sequence generation, has been the norm for training deep neural networks for years. Although being a widely-adopted approach, using cross entropy as th…
View article: Searching Toward Pareto-Optimal Device-Aware Neural Architectures
Searching Toward Pareto-Optimal Device-Aware Neural Architectures Open
Recent breakthroughs in Neural Architectural Search (NAS) have achieved state-of-the-art performance in many tasks such as image classification and language understanding. However, most existing works only optimize for model accuracy and l…
View article: MONAS: Multi-Objective Neural Architecture Search using Reinforcement Learning
MONAS: Multi-Objective Neural Architecture Search using Reinforcement Learning Open
Recent studies on neural architecture search have shown that automatically designed neural networks perform as good as expert-crafted architectures. While most existing works aim at finding architectures that optimize the prediction accura…
View article: DC-Prophet: Predicting Catastrophic Machine Failures in DataCenters
DC-Prophet: Predicting Catastrophic Machine Failures in DataCenters Open
When will a server fail catastrophically in an industrial datacenter? Is it possible to forecast these failures so preventive actions can be taken to increase the reliability of a datacenter? To answer these questions, we have studied what…
View article: In-Situ Rs and Improvement in Thermal Stability of Nickel Silicides Using Different Interlayer Films
In-Situ Rs and Improvement in Thermal Stability of Nickel Silicides Using Different Interlayer Films Open
Different interlayer films (Mo, Ru, Ta, Ti and Zn) were proposed to reduce sheet resistance (Rs) and improve thermal stability for nickel (Ni) silicide formation. It was found that the Zn and Mo interlayers reduced nickel silicide Rs much …