Published Paper Details:

This study explores the implementation of an Adaptive In-Loop Filter (AILF) for High Efficiency Video Coding (HEVC) utilizing deep learning technique. The need for effective compression techniques that preserve excellent visual quality has grown as video content continues to spread. Even though the traditional in-loop filters perform well, they frequently have difficulties maximizing performance in a variety of video scenarios and environments. According to the properties of the video being analyzed, this study suggests an AILF that includes Convolutional Gated Recurrent Unit (ConvGRU), a type of Recurrent Neural Network typically involves enhancing reconstructed frames by exploiting temporal dependencies across frames. In addition to enhancing reconstructed frames, the AILF performs better than traditional techniques in terms of Structural Similarity Index (SSIM) and Peak Signal-to-Noise Ratio (PSNR), indicating its potential for practical uses in broadcasting and video streaming. By demonstrating how well deep learning techniques can be integrated into video processing tasks, this work adds to the continuous developments in video coding technology.