- talks - lists.cs.princeton.edu

Christopher Hodsdon will present his general exam on Tuesday, January 22, 2019 at 3pm in CS 302.
by Nicki Gotsis 16 Jan '19

16 Jan '19

Christopher Hodsdon will present his general exam on Tuesday, January 22, 2019 at 3pm in CS 302. The members of his committee are as follows: Wyatt Lloyd (Adviser), Michael Freedman, and Amit Levy. Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. Title: "Holeless, Fault Tolerant, and Scalable Sequencing" Abstract: Some recent distributed systems use a monolithic sequencer, a single server handing out sequence numbers, as a building block for ordering operations in the system. Monolithic sequencers can simplify the design of distributed systems by replacing complex distributed ordering protocols, such as two-phase locking and Paxos, with a single round trip to the sequencer. Sequencers can also improve system throughput by ordering operations before they are executed, reducing contention. However, existing monolithic sequencers have two limitations. First, they allow holes in the sequence, which makes it difficult to build correct and efficient systems. Second, they impose a scalability ceiling on the system. We present CATS, a sequencing subsystem that overcomes these limitations while still providing the simplicity and throughput benefits of a monolithic sequencer. CATS provides a holeless sequence with a specialized reliable transport protocol, a sequencer recovery protocol, and a replicated layer of proxies interposed between clients and the sequencer. CATS also leverages the replicated layer of proxies to provide infinitely scalable throughput through two complementary scaling mechanisms. ------- Reading List: Balakrishnan, Mahesh, Dahlia Malkhi, Vijayan Prabhakaran, Ted Wobber, Michael Wei, and John D. Davis. "Corfu: A shared log design for flash clusters." In Proceedings of the 9th USENIX conference on Networked Systems Design and Implementation, pp. 1-1. USENIX Association, 2012. Wei, Michael, Amy Tai, Christopher J. Rossbach, Ittai Abraham, Maithem Munshed, Medhavi Dhawan, Jim Stabile, Udi Wieder, Scott Fritchie, Steven Swanson, Michael J. Freedman, and Dahlia Malkhi. "vCorfu: A Cloud-Scale Object Store on a Shared Log." In NSDI, pp. 35-49. 2017. Li, Jialin, Ellis Michael, and Dan RK Ports. "Eris: Coordination-free consistent transactions using in-network concurrency control." In Proceedings of the 26th Symposium on Operating Systems Principles, pp. 104-120. ACM, 2017. Li, Jialin, Ellis Michael, Naveen Kr Sharma, Adriana Szekeres, and Dan RK Ports. "Just Say NO to Paxos Overhead: Replacing Consensus with Network Ordering." In OSDI, pp. 467-483. 2016. Lamport, Leslie. "The part-time parliament." ACM Transactions on Computer Systems (TOCS) 16, no. 2 (1998): 133-169. Liskov, Barbara, and James Cowling. "Viewstamped replication revisited." (2012). Moraru, Iulian, David G. Andersen, and Michael Kaminsky. "Egalitarian paxos." In ACM Symposium on Operating Systems Principles. 2012. Kapritsos, Manos, Yang Wang, Vivien Quema, Allen Clement, Lorenzo Alvisi, and Mike Dahlin. "All about Eve: Execute-Verify Replication for Multi-Core Servers." In OSDI, vol. 12, pp. 237-250. 2012. Schneider, Fred B. "Implementing fault-tolerant services using the state machine approach: A tutorial." ACM Computing Surveys (CSUR) 22, no. 4 (1990): 299-319. Lamport, Leslie. "Time, clocks, and the ordering of events in a distributed system." Communications of the ACM 21, no. 7 (1978): 558-565. Textbook: Tanenbaum, Andrew S., and Maarten Van Steen. Distributed systems: principles and paradigms. Prentice-Hall, 2007.

1 0

Khiem's General Exam will take place on Tuesday, January 22, 2019 in CS 105 at 12pm.
by Nicki Gotsis 16 Jan '19

16 Jan '19

BQ_BEGIN Khiem's General Exam will take place on Tuesday, January 22, 2019 in CS 105 at 12pm. The members of his committee are as follows: Wyatt Lloyd (Adviser), Michael Freedman, and Amit Levy. BQ_BEGIN Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. BQ_END Title: K2: Lower Latency for Partial Replication Across Many Datacenters Abstract: The infrastructure available to large-scale web services now spans dozens of geographically dispersed datacenters. Utilizing this large number of datacenters to scale storage capacity requires a backend storage system to partially replicate data, i.e., to store only a subset of data in each datacenter. Partial replication, however, sacrifices the low latency benefits of having many datacenters: accessing data that is not replicated in a local datacenter incurs higher delay from contacting far-away datacenters. K2, a causally-consistent data store with read-only transactions and write-only transactions, shows it is possible for a partially-replicated storage system to provide local datacenter latency in many settings. This is powered by our novel replication approach, MetaDC, and a novel read-only transaction algorithm. With only simple writes, MetaDC provides the best possible worst-case latency for read-only transactions in K2: a single round trip to remote datacenters. The novel read-only transaction algorithm allows many read-only transactions to execute entirely in their local datacenters by exploiting a small data cache. Our evaluation shows that K2 achieves much lower latency than a baseline that directly adapts causal consistency for partial replication. Reading List Textbook: TANENBAUM, A. S., AND vAN STEEN, M. Distributed Systems: Principles and Paradigms (2nd Edition).Pearson Education, 2006. Paper list: AHAMAD, M., NEIGER, G., KOHLI, P., BURNS, J., AND HUTTO, P. Causal Memory: Definitions, Implementation, and Programming. Distributed Computing 9, 1 (1995). ANNAMALAI, M., RAVICHANDRAN, K., SRINIVAS, H., ZINKOVSKY, I., PAN, L., SAVOR, T., NAGLE, D., AND STUMM, M. Sharding the Shards: Managing Datastore Locality at Scale with Akkio. In Proc. OSDI (2018). BELARAMANI, N., DAHLIN, M., GAO, L., NAYATE, A., VENKATARAMANI, A., YALAGANDULA, P., AND ZHENG, J. PRACTI replication. In Proc. NSDI (2006). CORBETT, J. C., DEAN, J., EPSTEIN, M., FIKES, A., FROST, C., FURMAN, J., GHEMAWAT, S., GUBAREV, A., HEISER, C., HOCHSCHILD, P., HSIEH, W., KANTHAK, S., KOGAN, E., LI, H., LLOYD, A., MELNIK, S., MWAURA, D., NAGLE, D., QUINLAN, S., RAO, R., ROLIG, L., SAITO, Y., SZYMANIAK, M., TAYLOR, C., WANG, R., ANDWOODFORD, D. Spanner: Google’s Globally-Distributed Database. In Proc. OSDI (2012). DECANDIA, G., HASTORUN, D., JAMPANI, M., KAKULAPATI, G., LAKSHMAN, A., PILCHIN, A., SIVASUBRAMANIAN, S., VOSSHALL, P., AND VOGELS , W. Dynamo: Amazon’s highly available key-value store. In Proc. SOSP (2007). LAMPORT, L. Time, clocks, and the ordering of events in a distributed system. Comm. ACM 21, 7 (1978). LAMPORT, L. The Part-Time Parliament. ACM TOCS 16, 2 (1998). LLOYD, W., FREEDMAN, M. J., KAMINSKY, M., AND ANDERSEN, D. G. Don’t Settle for Eventual: Scalable Causal Consistency for Wide-Area Storage with COPS. In Proc. SOSP (2011). LLOYD, W., FREEDMAN, M. J., KAMINSKY, M., AND ANDERSEN, D. G. Stronger Semantics for Low- Latency Geo-Replicated Storage. In Proc. NSDI (2013). MAHMOOD, T., NARAYANAN, S. P., RAO, S., VIJAYKUMAR, T. N., AND THOTTETHODI, M. Karma: Cost-effective geo-replicated cloud storage with dynamic enforcement of causal consistency. IEEE TCC (2018). PETERSEN, K., SPREITZER, M., TERRY, D., THEIMER, M., AND DEMERS, A. Flexible update propagation for weakly consistent replication. In Proc.SOSP (1997). ZAWIRSKI, M., PREGUIC ̧A, N., DUARTE, S., BIENIUSA, A., BALEGAS, V., AND SHAPIRO, M. Write Fast, Read in the Past: Causal Consistency for Client-Side Applications. In ACM/IFIP/USENIX International Middleware Conference (2015). BQ_END

1 0

2-3pm Wed Jan 23rd Larry Peterson distinguished colloquium in CS 105
by Jennifer Rexford 14 Jan '19

14 Jan '19

On Wednesday January 23rd, Larry Peterson, emeritus faculty and former chair of computer science at Princeton, will give a distinguished colloquium talk. Note the unusual 2pm start time. Title: Democratizing the Network Edge Speaker: Larry Peterson, Princeton University and Open Networking Foundation Date/time: 2-3pm Wed Jan 23 Location: CS 105 (the “small auditorium”) Abstract: The network edge—where the access networks that connect homes, businesses, and mobile users to the Internet are implemented—is in the midst of a transformation. Network operators (Telcos and CableCos) are transitioning from closed and proprietary hardware to disaggregated and virtualized software running on white-box servers, switches, and access devices. This transformation is partly motivated by the CAPEX savings that come from replacing purpose-built appliances with commodity hardware, but it is mostly driven by the need to increase feature velocity through the softwarization of the access network. The goal is to enable new classes of edge services—e.g., Autonomous Vehicles, Internet-of-Things (IoT), Immersive User Interfaces—that benefit from low latency connectivity. This is all part of the growing trend to move functionality to the network edge. This transformation faces two technical challenges. The first is to disaggregate and virtualize existing hardware appliances. This is done, for example, by applying SDN principles (decoupling the network control and data planes) and by breaking monolithic applications into a set of micro-services. The second is to integrate the resulting disaggregated components back into a functioning system, packaged with sufficient configuration and life cycle management tools to make it deployable in a production environment. This talk describes CORD, a community-based open source effort to address these challenges. CORD is being deployed in Telco field trials, but more interestingly, it lowers the barrier for anyone (not just global carriers) to deploy edge solutions that incorporate the latest 5G (wireless) and XGS-PON (fiber) access technologies. The talk concludes with some remarks on the opportunity this democratization of the network edge opens for the research community. Bio: Larry Peterson is the Robert E. Kahn Professor of Computer Science, Emeritus at Princeton University, where he served as Chair from 2003-2009. He is a co-author of the best selling networking textbook Computer Networks: A Systems Approach (5e). His research focuses on the design, implementation, and deployment of Internet-scale distributed systems, including the widely used PlanetLab and MeasurementLab platforms. He is currently working on a new access edge cloud called CORD, an open source project of the Open Networking Foundation.

1 0

Santiago Cuellar will present his Pre FPO on Wednesday, January 23, 2019 at 10am in CS 402.
by Nicki Gotsis 11 Jan '19

11 Jan '19

Santiago Cuellar will present his Pre FPO on Wednesday, January 23, 2019 at 10am in CS 402. The members of his committee are as follows: Andrew Appel (adviser), Readers: Chung-Kil (Seoul National University) and Zak Kincaid; Non readers: Lennart Beringer and Aarti Gupta All are welcome to attend. Abstract follows below. Concurrent Permission Machine for modular proofs of optimizing compilers with shared memory concurrency. Optimizing compilers change a program based on a formal analysis of its code and modern processors further rearrange the program order. It is hard to reason about such transformations, which makes them a source of bugs, particularly for concurrent shared-memory programs where the order of execution is critical. On the other hand, programmers should reason about their program in the source language, which abstracts such low level details. We present the Concurrent Permission Machine (CPM), a semantic model for shared-memory concurrent programs, which is: (1) sound for high-order Concurrent Separation Logic, (2) convenient to reason about compiler correctness, and (3) useful for proving reduction theorems on weak memory models. The key feature of the CPM is that it exploits the fact that correct shared-memory programs are permission coherent: threads have (at any given time) noncompeting permission to access memory, and their load/store operations respect those permissions. Compilers are often written with sequential code in mind, and proving the correctness of those compilers is hard enough without concurrency. Indeed, the machine-checked proof of correctness for the CompCert C compiler was a major advance in the field. Using the CPM to conveniently distinguish sequential execution from concurrent interactions, I show how to reuse the (sequential) CompCert proof, without major changes, to guarantee a stronger concurrent-permission-aware notion of correctness.

1 0

Linguang Zhang will present his Pre FPO "Towards a Robust Local Feature Pipeline" on Tuesday, January 15, 2019 at 1pm in CS 302
by Nicki Gotsis 11 Jan '19

11 Jan '19

Linguang Zhang will present his Pre FPO on Tuesday, January 15, 2019 at 1pm in CS 302. The members of his committee are as follows: Szymon Rusinkiewicz (advisor), Jia Deng, Adam Finkelstein, Thomas Funkhouser, and Olga Russakovsky. All are welcome to attend. Abstract follows below. Title: Towards a Robust Local Feature Pipeline Many computer vision applications including image matching, image-based reconstruction and localization rely on extracting and matching robust local features. A typical local feature pipeline first detects repeatable keypoints in the image (i.e., keypoint detector), and then computes a short vector to uniquely describe each keypoint (i.e., feature descriptor). Both the keypoint detector and the feature descriptor are conventionally hand-crafted based on what is intuitive to the designer. For example, corners or blobs are popular choices of keypoints, and the image gradient is a useful clue for descriptors. However, it is often difficult to define these principles to accommodate various applications. In this thesis, we study data-driven approaches which can more easily tailor the local feature pipeline for target applications. We start with a mobile robotics application that leverages local features extracted from ground texture images to achieve high-precision global localization. The second part of the thesis addresses the problem that existing keypoint detectors that are optimized for natural images suffer from sub-optimal performance on texture images. We therefore learn a keypoint detector specifically for each type texture using a deep neural network. Our detector automatically learns to identify keypoints that are distinctive in the target texture rather than relying on a set of pre-defined rules. Finally, we focus on a non-parametric approach for learning feature descriptors. Many well-performing local feature descriptors are trained using a triplet loss that includes a tunable margin, which limits its ability to generalize to other types of data and problems. We propose to replace the hard margin with a soft margin that self-tunes as learning progresses. To summarize, we first demonstrate through a novel visual-based localization system where a customized local feature pipeline is critical. Then, we tackle both the keypoint detector and the feature descriptor with generalizable data-driven approaches.

1 0

Borislav Hristov will present his Pre-FPO on Wednesday, Jan 9th, 2019 at 11am in LSI 200.
by Nicki Gotsis 08 Jan '19

08 Jan '19

Borislav Hristov will present his Pre-FPO on Wednesday, Jan 9th, 2019 at 11am in LSI 200. The members of his committee are as follows: Mona Singh (advisor), Barbara Engelhardt, Olga Troyanskaya, Ben Raphael, Bernard Chazelle Everyone is invited to attend his talk. Please see abstract below. Dramatically decreased cost of sequencing has enabled us to sequence the genomes of hundreds of individuals with various diseases. Yet, pinpointing the gene variants responsible for the development of heterogeneous diseases remains a particularly hard task because the same phenotypic outcome (disease) can result from a myriad of combinations of different alterations across the genome. In this talk, I will describe two projects that further our ability to computationally highlight potential disease causing genes by examining the disease genomes in the context of biological networks. First, I will present a novel network-based approach to uncover cancer-relevant genes from large-scale cancer genomics data. My approach tackles cancer mutational heterogeneity by utilizing per-individual mutational profiles. My method is based on the expectation that if a pathway is relevant for cancer, then (1) many individuals will have a somatic mutation within one of the genes comprising the pathway and (2) the genes comprising the pathway will interact with each other and together form a small connected subcomponent within the larger network. I will provide an intuitive formulation relying on balancing the size of the connected subgraph with covering many patients. I will describe a machine learning-like schema for selecting the value of the single required parameter and both an integer linear programming framework and a fast heuristic for optimizing the objective function. I will demonstrate the method’s outstanding ability to identify cancer-relevant genes, especially those mutated at very low rates. Second, I will describe a general computational framework that uses prior knowledge of disease associated genes to guide a network-based search for novel ones based on newly acquired information. While many methods that directly consider either the prior knowledge or the newly acquired information to uncover disease-relevant genes have been developed, I develop a new network-based approach that combines these two distinct types of information. I will show how such a framework built upon graph kernels and guided random walks has the power to uncover new disease associated genes by applying it to different heterogeneous diseases, including macular degeneration and eight cancer types.

1 0

Fwd: TOMORROW, January 9, 2019 at 4:30pm in 101 Sherrerd Hall - Jason Lee, University of Southern California
by Jen Rexford 08 Jan '19

08 Jan '19

> > Speaker: > Jason Lee, Marshall School of Business, University of Southern California > Title: > On the Foundations of Deep Learning: SGD, Overparametrization, and Generalization > Day: > Wednesday, January 9, 2019 > Time: > 4:30 pm > Room: > 101 Sherrerd Hall > > > Abstract: > We provide new results on the effectiveness of SGD and overparametrization in deep learning. > a) SGD: We show that SGD converges to stationary points for general nonsmooth , nonconvex functions, and that stochastic subgradients can be efficiently computed via Automatic Differentiation. For smooth functions, we show that gradient descent, coordinate descent, ADMM, and many other algorithms, avoid saddle points and converge to local minimizers. For a large family of problems including matrix completion and shallow ReLU networks, this guarantees that gradient descent converges to a global minimum. > b) Overparametrization: We show that gradient descent finds global minimizers of the training loss of overparametrized deep networks in polynomial time. > c) Generalization: For general neural networks, we establish a margin-based theory. The minimizer of the cross-entropy loss with weak regularization is a max-margin predictor, and enjoys stronger generalization guarantees as the amount of overparametrization increases. > d) Algorithmic and Implicit Regularization: We analyze the implicit regularization effects of various optimization algorithms on overparametrized networks. In particular we prove that for least squares with mirror descent, the algorithm converges to the closest solution in terms of the bregman divergence. For linearly separable classification problems, we prove that the steepest descent with respect to a norm solves SVM with respect to the same norm. For over-parametrized non-convex problems such as matrix sensing or neural net with quadratic activation, we prove that gradient descent converges to the minimum nuclear norm solution, which allows for both meaningful optimization and generalization guarantees. > > > > Bio: > Jason Lee is an assistant professor in Data Sciences and Operations at the University of Southern California. Prior to that, he was a postdoctoral researcher at UC Berkeley working with Michael Jordan. Jason received his PhD at Stanford University advised by Trevor Hastie and Jonathan Taylor. His research interests are in statistics, machine learning, and optimization. Lately, he has worked on high dimensional statistical inference, analysis of non-convex optimization algorithms, and theory for deep learning. >

1 0

Minsung Kim will present his General Exam on January 14, 2019 at 3pm in CS 402.
by Nicki Gotsis 07 Jan '19

07 Jan '19

Minsung Kim will present his General Exam on January 14, 2019 at 3pm in CS 402. The members of his committee are as follows: Kyle Jamieson (adviser), Jennifer Rexford, and Andrew Houck (ELE). Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. Abstract : User demand for increasing amounts of wireless capacity continues to outpace supply, and so to meet this demand, significant progress has been made in new MIMO wireless physical layer techniques. Higher-performance systems now remain impractical largely because their algorithms are extremely computationally demanding. For optimal performance, an amount of computation that increases at an exponential rate both with the number of users and with the data rate of each user is often required. The base station's computational capacity is thus becoming the limiting factor on wireless capacity. QuAMax is the first large MIMO wireless base station design that addresses this issue with a fundamentally different type of computation called quantum annealing. QuAMax solves computationally-demanding MIMO maximum likelihood signal detector. We argue that using quantum computing assisted technology in the future could hold the potential to overcome computational limits of digital electronics on the number of mobile users that a MIMO base station with a particular number of antennas could effectively serve. We have implemented QuAMax on the 2,031 qubit D-Wave 2000Q quantum annealer, the state-of-art of the field. Our experimental results evaluate that implementation on real and synthetic MIMO channel traces, showing under what conditions the quantum annealer can maintain optimal maximum likelihood performance for 30 user, 30 AP antenna QPSK communication, or 60 user, 60 AP antenna BPSK communication. Our work paves the way for quantum hardware and software to significantly expand the performance envelope of Massive MIMO. Reading list : [1] King, Andrew D and McGeoch, Catherine C. "Algorithm engineering for a quantum annealing platform." arXiv preprint arXiv:1410.2628 (2014) [2] E. D. Dahl. "Programming with D-Wave: Map Coloring Problem." D-Wave whitepaper (2013) [3] Perdomo-Ortiz, Alejandro, et al. "A performance estimator for quantum annealers: Gauge selection and parameter setting." arXiv preprint arXiv:1503.01083 (2015). [4] King, James, et al. "Benchmarking a quantum annealing processor with the time-to-target metric." arXiv preprint arXiv:1508.05087 (2015). [5] Venturelli, Davide, and Alexei Kondratyev. "Reverse quantum annealing approach to portfolio optimization problems." arXiv preprint arXiv:1810.08584 (2018). [6] Halperin, Daniel, et al. "802.11 with multiple antennas for dummies." ACM SIGCOMM Computer Communication Review 40.1 (2010) [7] Chan, Albert M., and Inkyu Lee. "A new reduced-complexity sphere decoder for multiple antenna systems." Communications, 2002. ICC 2002. IEEE International Conference on . Vol. 1. IEEE, 2002. [8] Nikitopoulos, Konstantinos, et al. "Geosphere: Consistently turning MIMO capacity into throughput." ACM SIGCOMM Computer Communication Review . Vol. 44. No. 4. ACM, 2014. [9] Barbero, Luis G., and John S. Thompson. "Fixing the complexity of the sphere decoder for MIMO detection." IEEE Transactions on Wireless Communications 7.6 (2008). [10] Abari, Omid, et al. "Enabling High-Quality Untethered Virtual Reality." NSDI . 2017.

1 0

Zhuqi Li will present his General Exam on Wednesday, January 9, 2019 at 10am in CS 402.
by Nicki Gotsis 02 Jan '19

02 Jan '19

Zhuqi Li will present his General Exam on Wednesday, January 9, 2019 at 10am in CS 402. The members of his committee are Kyle Jamieson (adviser), Jennifer Rexford, and Wyatt Lloyd. Everyone is invited to attend his talk, and those faculty wishing to remain for the oral exam following are welcome to do so. His abstract and reading list follow below. Abstract: Conventional thinking treats the wireless channel as a given constraint. Therefore, wireless network designs to date center on the problem of the endpoint optimization that best utilizes the channel, for example, via rate and power control at the transmitter or sophisticated decoding mechanisms at the receiver. We instead explore whether it is possible to reconfigure the environment itself to facilitate wireless communication. In this work, we instrument the environment with a large array of inexpensive antennas (LAIA) and design algorithms to configure them in real time. Our system achieves this level of programmability through rapid adjustments of an on-board phase shifter in each LAIA device. We design a channel decomposition algorithm to quickly estimate the wireless channel due to the environment alone, which leads us to a process to align the phases of the array elements. Variations of our core algorithm can then optimize wireless channels on the fly for single- and multi-antenna links, as well as nearby networks operating on adjacent frequency bands. We design and deploy a 36-element passive array in a real indoor home environment. Experiments with this prototype show that, by reconfiguring the wireless environment, we can achieve a 24% TCP throughput improvement on average and a median improvement of 51.4% in Shannon capacity over the baseline single-antenna links. Over the baseline multi-antenna links, LAIA achieves an improvement of 12.23% to 18.95% in Shannon capacity. The book I am going to use is "Computer Networks: A Systems Approach 4th Edition" by Larry L. Peterson, Bruce S. Davie. The list of papers are: [1] Cerf, Vinton, and Robert Kahn. "A protocol for packet network intercommunication." IEEE Transactions on communications (1974) [2] Kung, Hsiang-Tsung, and John T. Robinson. "On optimistic methods for concurrency control." ACM Transactions on Database Systems (TODS) (1981) [3] Bharghavan, Vaduvur, et al. "MACAW: a media access protocol for wireless LAN's." ACM SIGCOMM Computer Communication Review (1994) [4] Perry, Jonathan, et al. "Spinal codes." ACM SIGCOMM 2012 [5] Miu, Allen, Hari Balakrishnan, and Can Emre Koksal. "Improving loss resilience with multi-radio diversity in wireless networks." ACM MobiCom 2005. [6] Gollakota, Shyamnath, and Dina Katabi. Zigzag decoding: combating hidden terminals in wireless networks. ACM SIGCOMM 2008. [7] Liu, Vincent, et al. "Ambient backscatter: wireless communication out of thin air." NSDI. 2013. [8] Qiao, Yue, et al. "PhyCloak: Obfuscating Sensing from Communication Signals." NSDI. 2016. [9] Abari, Omid, et al. "Enabling High-Quality Untethered Virtual Reality." NSDI. 2017. [10] Xiong, Xi, et al. "Customizing indoor wireless coverage via 3D-fabricated reflectors." ACM BuildSys 2017.

1 0

Elena Balashova will present her Pre FPO "Structure-Aware Shape Synthesis" on December 13, 2018 at 12:30pm in CS 105.
by Nicki Gotsis 07 Dec '18

07 Dec '18

Elena Balashova will present her Pre FPO "Structure-Aware Shape Synthesis" on December 13, 2018 at 12:30pm in CS 105. The members of her committee are as follows: Tom Funkhouser (adviser), Szymon Rusinkiewicz, Vivek Singh (Siemens), Adam Finkelstein, Olga Russakovsky All are invited to attend. Abstract follows below. Abstract: Shape synthesis is an important area of computer vision and graphics that concerns creation of new shapes and reconstruction from partial data. Its goal is to learn a model that can generate shapes within an object category suitable for novel shape creation, interpolation, completion, editing, and other geometric modeling applications. Existing tools learn shape properties from large collections of shapes. Although these methods have been very successful at learning how to synthesize the coarse shapes of objects in categories with highly diverse shapes, they have not always produced examples that reconstruct important structural elements of a shape. In this talk, I describe how structure can be incorporated into the synthesis process, and how it can be used to improve generative models. I discuss how the specific properties of the data influence the effect and design of chosen structure. First, I introduce a template-defined skeleton structure for learning a part-aware generative model in typography, where the shapes have a known structure and can be explained by a small number of templates. Next, I describe a semantic landmark-based structure and how it can be used to improve a generative model of examples with extremely varied topology by means of a geometric shape-structure consistency loss. Finally, I present a scenario of noisy archaeological wall painting (fresco) reconstruction from eroded fragments, where there is no well-defined structure and exponentially many arrangement possibilities – in this case, I show that structure can inferred and assembled through consistency of arrangement

1 0