MobileNetV2
3 articles · search the full text for this term
-
Real-Time Deepfake Detection in Video Conferencing Systems
Abstract: Deepfake technology presents non-exemplary threats to video conferencing platforms, enabling advanced fraud, impression and misinformation campaigns worth billions annually. Current detection methods either exhibit latencies exceeding 100ms or rely on server-side cloud processing, raising privacy concerns. This paper presents DeepConfGuard, a lightweight hybrid architecture combining MobileNetV2 for spatial feature extraction, a bidirectional LSTM with attention for temporal modelling, and EfficientNetV2 for refinement. It reaches 94.8% accuracy with 85 ms end‑to‑end …
Published in International Journal of Electronics Automation · Vol. 4, Issue 2, 2026 Read article
-
CNN-Based Wound Segmentation: A Review of Models and Performance Evaluation
Abstract: Deep learning, particularly convolutional neural networks (CNNs), has altered medical image processing by automating and precisely segmenting complex medical pictures. Wound segmentation, a critical application in automated wound assessment, is essential for wound size estimation, classification, and healing progress monitoring. This study presents a comprehensive review of CNN-based wound segmentation models, focusing on their architectures, methodologies, and performance on diverse datasets. Four deep learning models, including two U-Net variants (5-layer …
Published in Current Trends in Signal Processing · Vol. 15, Issue 1, 2025 · pp. 33–46 Read article
-
A Comparative Analysis of Machine Learning Techniques for Fruit Defect Detection Systems
Abstract: With evolving technologies in machine learning, significant advancements have been made in the livestock industry, helping to reduce waste, increase yield, achieve cost savings, and improve competitiveness in the marketplace. Fruit defect detection models support precision agriculture by providing valuable data for decision-making and enhancing overall efficiency through automated inspection processes. This study implements and comparatively evaluates machine learning models including MobileNetV2, a custom-designed convolutional neural network (CNN) model, ResNet50, …
Published in Journal of Image Processing & Pattern Recognition Progress · Vol. 11, Issue 3, 2024 · pp. 37–47 Read article