Journal clubs are held in the STIM Laboratory (W309). For each session, a participating student or researcher will select a relevant paper, send that paper to the group, and present it for discussion. The number of sessions will depend on the number of participants, with the goal of each student presenting 1-2 papers per year.
Microscopy and Microscope Image Processing
Mondays at 11am - 12pm (STIM Laboratory - Engineering Building 2, W309)
This journal club focuses on the hardware and algorithms for acquiring and processing high-resolution images. Based on the interests of the labs involved, presented papers will generally focus on large biomedical images collected using optical microscopy, including bright-field, fluorescence, and infrared imaging.
Relevant topics include: new imaging techniques, three-dimensional imaging, image processing, segmentation, hyperspectral image analysis, noise removal, and reconstruction
Sample Papers / Journals
"Expansion microscopy with conventional antibodies and fluorescent proteins."
Nature Methods (2016)
Chozinski, Halpern, Okawa, Kim, Tremel, Wong, and Vaughan
"Fast 3D visualization of endogenous brain signals with high-sensitivity laser scanning photothermal microscopy."
Biomedical Optics Express (2016)
Miyazaki, Iida, Tanaka, Hayashi-Takagi, Kasai, Okabe, and Kobayashi
"Hermite Snakes with Control of Tangents."
IEEE Transactions on Image Processing (2016)
Uhlmann, Fageot, and Unser
"NeuroBlocks: Visual Tracking of Segmentation and Proof-Reading for Large Connectomics Projects"
IEEE Transactions on Visualization and Computer Graphics (2016)
Ai-Awami, Beyer, Haehn, Kasthuri, Lichtman, Pfister, and Hadwiger
Wednesdays at 4pm - 5pm (STIM Laboratory - Engineering Building 2, W309)
This journal club will focus on high-performance heterogeneous computing, with a particular focus on highly parallel architectures such as GPUs and the Xeon Phi. We expect the audience to be primarily electrical engineers with some understanding of the underlying GPU architecture.
Relevant topics include: new GPU architectures, GPU-based algorithms, heterogeneous algorithms, CUDA
|5/18||Kedar Grama||Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks|
|Srijani Mukherjee||Edge-Preserving Image Denoising via Group Coordinate Descent on the GPU|
Sample Papers / Journals
"GeePS: Scalable deep learning on distributed GPUs with a GPU-specialized parameter server."
Proceedings of the Neural Information Professing Systems Conference (NIPS) (2016)
Ren, He, Girshick, and Sun
"GPU-Based Fuzzy C-Means Clustering Algorithm for Image Segmentation."
Almazrooie, Vadiveloo, and Abdullah