Deep Convolutional Neural Network
Deep convolutional neural networks (CNNs) are a class of artificial neural networks designed to process data with a grid-like topology, such as images and videos, excelling at tasks like image classification, object detection, and segmentation. Current research focuses on improving CNN architectures (e.g., exploring variations of ResNet, Inception, and efficientNet models), developing novel training techniques (like integer-only training and self-knowledge distillation), and addressing challenges such as imbalanced datasets and catastrophic forgetting in incremental learning. The widespread application of CNNs across diverse fields, from medical image analysis and autonomous driving to agricultural monitoring and remote sensing, highlights their significant impact on both scientific understanding and practical problem-solving.
Papers
SaiT: Sparse Vision Transformers through Adaptive Token Pruning
Ling Li, David Thorsley, Joseph Hassoun
RoHNAS: A Neural Architecture Search Framework with Conjoint Optimization for Adversarial Robustness and Hardware Efficiency of Convolutional and Capsule Networks
Alberto Marchisio, Vojtech Mrazek, Andrea Massa, Beatrice Bussolino, Maurizio Martina, Muhammad Shafique
Evaluation of importance estimators in deep learning classifiers for Computed Tomography
Lennart Brocki, Wistan Marchadour, Jonas Maison, Bogdan Badic, Panagiotis Papadimitroulas, Mathieu Hatt, Franck Vermet, Neo Christopher Chung
Convolutional Neural Networks Quantization with Attention
Binyi Wu, Bernd Waschneck, Christian Georg Mayr