RFC-Net: Learning High Resolution Global Features for Medical Image Segmentation on a Computational Budget (Student Abstract)
Keywords:Computer Vision, Machine Learning, AI Architectures, Applications Of AI
AbstractLearning High-Resolution representations is essential for semantic segmentation. Convolutional neural network (CNN) architectures with downstream and upstream propagation flow are popular for segmentation in medical diagnosis. However, due to performing spatial downsampling and upsampling in multiple stages, information loss is inexorable. On the contrary, connecting layers densely on high spatial resolution is computationally expensive. In this work, we devise a Loose Dense Connection Strategy to connect neurons in subsequent layers with reduced parameters. On top of that, using a m-way Tree structure for feature propagation we propose Receptive Field Chain Network (RFC-Net) that learns high-resolution global features on a compressed computational space. Our experiments demonstrates that RFC Net achieves state-of-the-art performance on Kvasir and CVC-ClinicDB benchmarks for Polyp segmentation. Our code is publicly available at github.com/sourajitcs/RFC-NetAAAI23.
How to Cite
Saha, S., Saha, S., Gani, M. O., Oates, T., & Chapman, D. (2023). RFC-Net: Learning High Resolution Global Features for Medical Image Segmentation on a Computational Budget (Student Abstract). Proceedings of the AAAI Conference on Artificial Intelligence, 37(13), 16314-16315. https://doi.org/10.1609/aaai.v37i13.27017
AAAI Student Abstract and Poster Program