PolypNextLSTM: A lightweight and fast polyp video segmentation network using ConvNext and ConvLSTM

arXiv:2402.11585v1 Announce Type: cross
Abstract: Commonly employed in polyp segmentation, single image UNet architectures lack the temporal insight clinicians gain from video data in diagnosing polyps. To mirror clinical practices more faithfully, our proposed solution, PolypNextLSTM, leverages video-based deep learning, harnessing temporal information for superior segmentation performance with the least parameter overhead, making it possibly suitable for edge devices. PolypNextLSTM employs a UNet-like structure with ConvNext-Tiny as its backbone, strategically omitting the last two layers to reduce parameter overhead. Our …

abstract architectures arxiv clinical clinicians cs.cv cs.lg data deep learning image information insight network performance practices segmentation solution temporal type unet video video data