Blind Quality Assessment

Blind image quality assessment (BIQA) focuses on automatically predicting the perceived quality of images and videos without needing a reference image, crucial for applications like surveillance and autonomous driving. Recent research emphasizes developing robust BIQA models for diverse image types, including low-light, blurred, and 3D point cloud data, often employing deep neural networks with architectures like convolutional neural networks and incorporating multimodal information (e.g., image and text) or frequency-domain analysis. These advancements improve the accuracy and efficiency of quality assessment, enabling better monitoring of image and video quality in various real-world applications and providing valuable insights into human visual perception.