News

Deep learning–enhanced Raman spectroscopy enables accurate identification of synthetic cannabinoids
Synthetic cannabinoids, a class of new psychoactive substances, have emerged as a significant public health and social stability threat due to their structural diversity, rapid iteration, and stronger ...
IEEE1d
Integrated Monitoring of Stress and Damage Using CNN Deep Learning of Electromechanical Impedance Responses of CSA Sensor - IEEE Xplore
A multitask 2D CNN model is designed for integrated monitoring stress and damage in concrete specimens utilizing the raw impedance signatures of capsule-like smart aggregated (CSA). The fundamental ...
IEEE Spectrum on MSN2d
New Machine Vision Is More Energy Efficient—and More Human
A I vision models have improved dramatically over the past decade. Yet these gains have led to neural networks which, though effective, don’t share many characteristics with human vision. For example, ...
Qonversations on MSN2d
How AI helped a couple conceive after 18 years of infertility
Eighteen years, multiple countries, and countless disappointments—that is the back‑story of a couple who, at last, are expecting their first child. Their turning‑point came not from another round of ...
IEEE2d
Achieving Multisite Generalization for CNN-Based Disease Diagnosis Models by Mitigating Shortcut Learning - IEEE Xplore
When a learned model has high accuracy under familiar settings (internal testing) and a big drop in accuracy under slightly different circumstances (external testing) we suspect it is using shortcuts ...
IEEE3d
An automated early ischemic stroke detection system using CNN deep learning algorithm - IEEE Xplore
Over the past few years, stroke has been among the top ten causes of death in Taiwan. Stroke symptoms belong to an emergency condition, the sooner the patient is treated, the more chance the patient ...
IEEE3d
Comparative Analysis of Pre-Trained CNN Backbones and Machine Learning Techniques for High Similarity in Circular Object Image Classification - IEEE Xplore
This research presents a comprehensive comparative analysis of various pre-trained backbone models and machine learning techniques for output layers in convolutional neural networks (CNNs) applied to ...
IEEE4d
Dynamic Semantic Compression for CNN Inference in Multi-Access Edge Computing: A Graph Reinforcement Learning-Based Autoencoder - IEEE Xplore
This paper studies the computational offloading of CNN inference in dynamic multi-access edge computing (MEC) networks. To address the uncertainties in communication time and edge servers’ available ...
IEEE4d
A Survey on Hyperparameter Optimization of Machine Learning Models - IEEE Xplore
Hyperparameters in machine learning are those variables that are set before the training process starts and regulate several aspects of the behavior of the learning algorithm. In contrast to model ...
IEEE4d
An Interpretable Brain Tumour Classification Approach using Machine Learning and CNN-based features - IEEE Xplore
Cancer, a disease that does not discriminate; impacting people from all walks of life. Numerous types of cancer affect humans, each with distinct characteristics and treatment approaches. One of the ...
IEEE5d
CNN Based Dense Underwater 3D Scene Reconstruction by Transfer Learning Using Bubble Database - IEEE Xplore
Dense 3D shape acquisition of swimming human or live fish is an important research topic for sports, biological science and so on. For this purpose, active stereo sensor is usually used in the air, ...
IEEE5d
Movements Classification of Multi-Channel sEMG Based on CNN and Stacking Ensemble Learning - IEEE Xplore
In recent years, the analysis of surface electromyography (sEMG) signals by feature engineering and machine learning has developed rapidly. However, when feature engineering is applied to feature ...