Advancing Weakly-supervised Audio-Visual Video Parsing via Segment-wise Pseudo Labeling We propose a new method for the audio-visual video parsing task, please refer to the [arxiv paper] for more details.