Image and Video Processing and Recognition Based on Artificial Intelligence (Volume II)

© 2022 by the authors; CC BY-NC-ND license

Finger-vein recognition; motion blur image restoration; modified DeblurGAN; CNN; deep learning; stress recognition; stress database; spatial attention; temporal attention; facial landmark; pose estimation; convolutional neural network; lightweight; knowledge distillation; convolutional neural network; histogram oriented graphic; multi-camera multi-object tracking; detection quality; traffic sign recognition; sliding windows; meta learning; deep learning; classification; construction sign detection; image synthesis; cut-and-paste; perspective transformation; deep learning; group-sparsity loss; temporal attention module; video-based pedestrian-attribute recognition; gait recognition; self-supervised learning; pose estimation; multi-task learning; weakly-supervised learning; biometrics; cross-sensor fingerprints; fingerprint enhancement; cGAN; adversarial learning; deep learning; video summarization; spatial–temporal features; cross-attention; adaptive video streaming; A-LSTM networks; bit rate measurement; client–server model; HTTP; reference metrics; video quality; deep learning; image deraining; neural network; vision transformer; generative adversarial network; computer vision; information retrieval; content-based video retrieval; anomaly detection; generation error; feature trajectory smoothness; surveillance video; facial expression recognition; spatial transformation network; attention mechanism; triplet loss function; intra-similarity problem; breast cancer; ultrasonography; deep learning; anomaly detection; autoencoder; brain image registration; generation adversarial network; deep learning; hand detection; hand classification; YOLO-family networks; convolutional neural networks (CNNs); egocentric vision; video summarization; unsupervised learning; reinforcement learning; piecewise linear interpolation