July 2020 - talks - lists.cs.princeton.edu

Sergiy Popovych will present his Pre FPO "Visual Feature Extraction with Unsupervised Metric Learning" on Friday August 7th, 2020 at 8:30pm ET via Zoom.
by Nicki Mahler 31 Jul '20

31 Jul '20

Sergiy Popovych will present his Pre FPO " Visual Feature Extraction with Unsupervised Metric Learning" on Friday August 7th, 2020 at 8:30pm ET via Zoom. Zoom Link: [ https://princeton.zoom.us/j/94304733448 | https://princeton.zoom.us/j/94304733448 ] The members of his committee are as follows: Sebastian Seung (Advisor); Readers: Sebastian Seung and Viren Jain (Google); Examiners: Stephan Saalfeld (Janelia), Yoram Singer, and Kai Li Please see abstract below. Abstract: In serial section electron microscopy (ssEM), a brain volume is cut into a series of ultrathin sections, each of which is imaged via electron microscopy. Then the 2D images are combined to create a 3D image stack, from which the connections between neurons can be reconstructed. Images obtained from serial section electron microscopy often contain distortions of both affine and non-affine nature. Before the 3D stack can be analyzed, these tissue distortions have to be corrected. The problem of correcting these distortions, also known as 3D image registration or stack alignment, poses significant challenges for ssEM neuron reconstruction. This thesis focuses on convolutional neural network (ConvNet) based method that can successfully align large stacks, even in presence of large discontinuous deformations. The key behind the proposed method is a metric learning based technique that learns to extract dense noiseless representation of ssEM images. In this technique, two ConvNets are trained simultaneously. The first ConvNet is a feature extractor that is applied to the images independently. The extracted features are passed to the second ConvNet, which predicts a corrective deformation field. The joint loss of the two networks is to minimize the mean square error between the extracted features after the corrective field is applied, while maximizing the difference between the features before the corrective field is applied. Special techniques are employed to encourage feature density and to avoid degenerate solutions. The proposed technique is evaluated on real world datasets and is shown to outperform other existing approaches. This thesis also presents an efficient distributed pipeline that can be used to efficiently apply the proposed method to multi-terabyte datasets.

1 0

CORRECTION: Uthsav Chitra will present his General Exam "Techniques for Analyzing Network Interactions " on July 29, 2020 at 3:30pm via Zoom.
by Nicki Mahler 23 Jul '20

23 Jul '20

Uthsav Chitra will present his General Exam "Techniques for Analyzing Network Interactions " on July 29, 2020 at 3:30pm via Zoom. Zoom address: [ https://princeton.zoom.us/j/244319523 ] [ https://princeton.zoom.us/j/98645694130 | https://princeton.zoom.us/j/98645694130 ] The members of his committee are as follows: Ben Raphael (adviser), Bernard Chazelle, and Christopher Musco (NYU) Link to Abstract and Reading list: [ https://docs.google.com/document/d/13hiLv3k2mssuZpeY50LpTnzHL-d9xmIU8UTDLD6… | https://docs.google.com/document/d/13hiLv3k2mssuZpeY50LpTnzHL-d9xmIU8UTDLD6… ] Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

Chitra Uthsav will present his General Exam "Techniques for Analyzing Network Interactions " on July 29, 2020 at 3:30pm via Zoom.
by Nicki Mahler 22 Jul '20

22 Jul '20

Chitra Uthsav will present his General Exam "Techniques for Analyzing Network Interactions " on July 29, 2020 at 3:30pm via Zoom. Zoom address: [ https://princeton.zoom.us/j/244319523 ] [ https://princeton.zoom.us/j/98645694130 | https://princeton.zoom.us/j/98645694130 ] The members of his committee are as follows: Ben Raphael (adviser), Bernard Chazelle, and Christopher Musco (NYU) Link to Abstract and Reading list: [ https://docs.google.com/document/d/13hiLv3k2mssuZpeY50LpTnzHL-d9xmIU8UTDLD6… | https://docs.google.com/document/d/13hiLv3k2mssuZpeY50LpTnzHL-d9xmIU8UTDLD6… ] Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 0

Sinong Geng will present his General Exam on June 17, 2020 at 3pm via Zoom.
by Nicki Mahler 03 Jul '20

03 Jul '20

Sinong Geng will present his General Exam on June 17, 2020 at 3pm via Zoom. Zoom link: [ https://princeton.zoom.us/j/369310330 | https://princeton.zoom.us/j/369310330 ] The members of his committee are as follows: Ronnie Sircar (adviser), Mete Soner (ORFE), Ryan Adams, and Karthik Narasimhan Link to Abstract & Reading List: [ https://drive.google.com/file/d/17qk8jPp680v4RezIphsU-9k4P_eR3rqr/view?usp=… | https://drive.google.com/file/d/17qk8jPp680v4RezIphsU-9k4P_eR3rqr/view?usp=… ] [ https://drive.google.com/file/d/1FWncLeSw5yCQIQCCsGhyQhWQtU3vuf7a/view?usp=… | https://drive.google.com/file/d/1FWncLeSw5yCQIQCCsGhyQhWQtU3vuf7a/view?usp=… ] Everyone is invited to attend the talk, and those faculty wishing to remain for the oral exam following are welcome to do so.

1 1

Jonathan Balkind will present his Pre-FPO "Open Source Frameworks for Enabling Full-Stack Hardware-Software Research" on Friday, June 10, 2020 at 9am via Zoom
by Nicki Mahler 02 Jul '20

02 Jul '20

Jonathan Balkind will present his Pre-FPO "Open Source Frameworks for Enabling Full-Stack Hardware-Software Research" on Friday, June 10, 2020 at 9am via Zoom. Zoom information: Time: Jul 10, 2020 09:00 AM Eastern Time (US and Canada) [ https://princeton.zoom.us/j/99108232595?pwd=eXBzL3ZMdGdPY2N6cTB0a3pKVHd5QT09 | https://princeton.zoom.us/j/99108232595?pwd=eXBzL3ZMdGdPY2N6cTB0a3pKVHd5QT09 ] Meeting ID: 991 0823 2595 Password: 765960 The members of his committee are as follows: David Wentzlaff (adviser); Readers: Christopher Batten (Cornell University) and Amit Levy; Examiners: Aarti Gupta and Margaret Martonosi. Abstract: Hardware-software research routinely starts with modelling. First order models can capture much of the desired system behaviour needed to begin evaluating architectural research ideas. However, when those ideas come closer to realisation in real chips or they focus on cross-layer optimisations, more precise tools must be brought to bear in evaluating them. This dissertation focuses on the development of two such tools, intended to enable research across the hardware-software stack. The tools described in this dissertation have been open-sourced in both their designs and characterisation in order to enable others to evaluate their research ideas with high fidelity. The first tool is OpenPiton, a full-stack manycore research platform and the world’s first open-source, general-purpose, multithreaded manycore processor. OpenPiton provides a toolbox for researchers to design new manycore processors incorporating their chosen research modifications. This toolbox includes scripts and tools necessary to produce a variety of processor configurations and take those instances all the way to realisation in real silicon chips. By embracing open-source design, the system supports modification at all levels of the stack, from underlying circuits to software running above, and as a result has been adopted in a variety of research areas across computing. The second tool presented in this dissertation is the Bring Your Own Core (BYOC) platform. BYOC is designed to enable heterogeneous-ISA systems research in architecture, systems software, and languages. BYOC brings together cores of multiple ISAs in a single, cache-coherent, manycore system and introduces a new interface for new cores to be rapidly connected to the system. This dissertation describes the integration of several new cores into OpenPiton to build this BYOC platform and the means taken to specifically support these cores existing together in a heterogeneous configuration. This dissertation also details a characterisation of an FPGA prototype named Dicho, built with a combination of an OpenSPARC T1 SPARC V9 core and an Ariane RISC-V core.

1 0