2nd Learning for Computational Imaging (LCI) Workshop @ ICCV 2021:
Sensing, Reconstruction, and Analysis
The ICCV Workshop on Learning for Computational Imaging (LCI) is a venue for presenting recent advances and trends in the emerging field of computational imaging, in which learning and computation are the major ingredients of the highly effective imaging systems. Given the rapidly growing interest in next-generation imaging systems and their combination with computer vision tasks, the LCI workshop welcomes recent work on novel imaging pipelines such as smart imaging system design, blind compressed sensing, and task-driven imaging systems. The scope covers research topics ranging from novel computational imaging pipeline, image and system modeling, theory, algorithms, applications in various imaging modalities, as well as industrial imaging applications.
Since data-driven and large-scale optimization methods are becoming more and more popular in the computational imaging community, a major focus of LCI will be on advanced models and machine learning techniques, including deep learning approaches, sparse and low-rank modeling, manifold learning, unrolled architectures, learning convolutional and tensor models, etc., which can be applied to enhance the effectiveness and efficiency of various imaging systems. LCI also promotes recent works on learning theory for computational imaging, including performance guarantee, convergence analysis, learning model analysis, etc., which are critical for reliable and interpretable imaging systems.
Last but not least, the scope of LCI includes research on the plethora of computational imaging applications, which include various imaging modalities (e.g., MRI, radar, lidar, microscopy, computational photography, optics, etc.). LCI will encourage close interaction between mathematical and applied computational imaging researchers and practitioners, and bring together experts in academia and industry working in imaging theory and applications, with focus on machine learning, data modeling, computer vision, signal processing, inverse problems, compressed sensing, optimization, neuroscience, learning and computation-driven hardware, and related areas.
This workshop is a half-day online event on Oct 17 afternoon, which will include invited talks, oral and spotlight presentations of accepted papers.
(New) The recordings of all sessions are available now. Check out our Program
Past LCI Workshops [ LCI @ ICCV 2019 ]
Our full workshop program is online
NEW - Extended Abstract Submission:
We introduce a new track of extended abstract submission.
The length of extended abstracts can be 1-4 pages, including all figures, tables, and references. The contents (including references) and length of the extended abstract should be sufficient for the submission to be properly evaluated.
We invite submissions of extended abstracts of
The review will be single-blind. Note that there will be NO published proceedings for extended abstracts. Authors of accepted abstracts can present their works in LCI.
Use the Extended Abstract track on CMT.
Novel Learning-Driven Computational Imaging Systems: Smart imaging and sensing systems, Learned data acquisition, Blind compressed sensing, Task-driven imaging systems, etc.
Learning-based Modeling and Algorithms for Imaging: Deep learning approaches and architectures, Sparse and low-rank modeling, Dictionary and transform learning, Manifold learning, Unrolled architectures, Graphical models, Tensor models, Online learning, Plug-and-play models, Bayesian methods, etc.
Learning Theory for Computational Imaging: Performance guarantees for learning-based methods, Convergence analysis of learning algorithms, Generative model recovery analysis, Analysis of deep architectures, Theory for large-scale and distributed algorithms, etc.
Computational Imaging Applications: Magnetic resonance imaging, Radar imaging, Lidar, Computed tomography, Microscopy, Ultrasound, Hyperspectral imaging, Hybrid imaging, Computational photography, Neuroimaging, Dynamic imaging, Super-resolution, Inpainting, and novel and extreme imaging modalities and applications.
Ethics and Social Impacts of Learning for Computational Imaging: Ethics and impacts of machine learning in applications; Privacy aspects (e.g., of using medical images or in security applications); Fairness questions; Ethical, moral, and legal implications in biomedical imaging, healthcare, etc.