News
Welcome to CoSIP – Phase 2!
The second phase of DFG SPP Compressed Sensing in Information Processing (CoSIP) has started from July 2018.
Project leaders are expected to inform their group members to subscribe to the mailing list by sending an email to cosip@math.tu-berlin.de. This concerns as well the current PhD-students who were already present in phase 1 if they are also active in phase 2.
Web-page:
The new web-page is now available at: https://www.ti.rwth-aachen.de/SPP1798II/
For any request concerning the web-page, please use the contact address cosip.phase2@ti.rwth-aachen.de.
Publication report:
Project leaders can update the publications for each project by sending the latest bibtex file of their publications to cosip.phase2@ti.rwth-aachen.de. The name of the bibtex file should be the project number without spaces and special characters (example: AB12341.bib for the project AB 123/4-1). The preprints should use the entry “unpublished”.
CoSIP retreat on compressed sensing and learning theory
CoSIP retreat on compressed sensing and learning theory was successfully held in Spa, Belgium on September 12-13. The retreat was composed of gradate students and senior researchers from the group of prof. Rodulf Mathar (RWTH-TI), prof. Holger Rauhut (RWTH-LS C) and prof. Massimo Fornasier (TUM). The retreat included following talks and follow-up discussions.
- Massimo Fornasier (TUM), Learning shallow neural nets from fewest samples
- Hans-Christian Jung (LS C), Recovery guarantees for 1-bit Compressed Sensing with Gaussian circulant matrices
- Emilio Balda (TI), An Information Theoretic Study on Learning of Artificial Neural Networks
- Presentation of DFG project CoCoMIMO followed by discussion
- Presentation of DFG project SCoSNeL followed by discussion
- Arash Behboodi (TI), Structured signal recovery using generative models
- Johannes Maly (TUM), Distributed Compressed Sensing with One-Bit Measurements
- Niklas Koep (TI), Group-Sparse Signal Recovery with Block-Diagonal Measurement Matrices
- Jannis Kurtz (LS C), Group sparse model selection
- Arya Bangun (TI), Coherence bounds for sensing matrix in spherical harmonics expansion
- Alexander Stollenwerk, Robust 1-Bit Compressed Sensing via Hinge Loss Minimization
- Ulrich Terstiege, Gradient flow for autoencoders
- Christian Kümmerle (TUM), Relax, but not too much: Algorithms for Low-Rank Models in Signal Processing
- Jonathan Fell (LS C), l1-analysis minimization