Abstract: Conventional deconvolution methods utilize hand-crafted image priors to constrain the optimization. While deep-learning-based methods have simplified the optimization by end-to-end training, ...
Abstract: Seismic deconvolution is essential for extracting layer information from noisy seismic data, but it is an ill-posed problem with nonunique solutions. Inspired by classical optimization ...
2023-01 [NEW:tada:] This repository now contains the code of the ICCV paper and the extra contents of the extended version, including: 2. SnowflakeNet: Point Cloud Completion by Snowflake Point ...
If you want to train the models with Shape Constraint, please install AlphaTransform. Download our simulated galaxy dataset from Google Drive. To train the models ...