Several Illinois CS Research Projects Result in Best Paper Awards

3/15/2021 3:26:10 PM Aaron Seidlitz, Illinois CS

As Illinois CS research continues to sustain high levels of success, one indicator is that the papers produced here earn best paper awards at some of the most prominent computing conferences.

From the faculty guiding these projects to the students learning what it takes to develop well-regarded research, these honors are representative of the effort and dedication given to research at Illinois CS.

Congratulations to each of the following projects and the authors involved.

Best Paper, IEEE International Conference on Data Mining 2020

Jiawei Han
Jiawei Han

Out of 930 submissions accepted at this conference in 2020, this paper from Illinois CS professor Jiawei Han’s research group earned Best Paper recognition.

First author Carl Yang received his PhD from Illinois CS in Nov. 2020 and now is an assistant professor at Emory University. The second author, Jiayu Zhang, received his MS from Illinois CS in May 2020 and is now a PhD student at University of Washington.

Together with Han, this group authored the paper titled “Co-Embedding Network Nodes and Hierarchical Labels with Taxonomy Based Generative Adversarial Networks.” In the paper’s abstract, the group describes their results as proven to improve “network embedding on traditional tasks of node classification and link prediction, as well as novel tasks like conditional proximity search and fine-grained taxonomy layout.”

Best Paper Award, IEEE International Conference on Acoustics, Speech and Signal Processing 2020

Paris Smaragdis
Paris Smaragdis

Illinois CS professor Paris Smaragdis teamed with Google DeepMind research scientist Po-Sen Huang (PhD ECE ’14), Electrical and Computer Engineering professor Mark Hasegawa-Johnson, and Indiana University professor Minje Kim (PhD CS ’16) on a Best Paper at 2020 IEEE International Conference on Acoustics, Speech and Signal Processing.

Their paper, titled "Joint Optimization of Masks and Deep Recurrent Neural Networks for Monaural Source Separation," will be honored at ICASSP 2021 in Toronto, Canada. Their findings showed performance improvement based on their exploration of “joint optimization of masking functions and deep recurrent neural networks for monaural source separation tasks, including speech separation, singing voice separation, and speech denoising.”


Best Paper Award, IEEE International Symposium on Multimedia 2020

Klara Nahrstedt
Klara Nahrstedt

Professor Klara Nahrstedt worked with a team of Illinois CS students, including Mingyuan Wu, Bo Check, and Eric Lee, to produce a paper titled: “SEAWARE: Semantic-Aware View Prediction System for 360-degree Video Streaming.”

The work was done jointly with University of Massachusetts Amherst faculty, and earned Best Paper at the IEEE International Symposium on Multimedia 2020. Its focus is on navigation of streaming video content for viewers who are watching 360 video with HMD (Head-Mounted Displays) using view and object predictions.

 

 

Best Student Paper Award, IEEE CLOUD 2020

Chen Li
Chen Li

Illinois CS PhD graduates Chen Li and Abdul Dakkak earned the IEEE Cloud Best Student Paper Award for the title piece, “The Design and Implementation of a Scalable DL Benchmarking Platform.”

The paper is the core of Chen’s thesis and included fellow authors Wen-mei Hwu, faculty from Electrical & Computer Engineering, and Jinjun Xiong of the IBM T. J. Watson Research Center. Together they produced a work that addresses the fast-paced Deep Learning (DL) landscape, which “lacks a DL benchmarking platform to facilitate evaluation and comparison of DL innovations.” The paper identifies 10 design features that are desirable within a DL benchmarking platform. 

Best Paper Award, IEEE International Parallel and Distributed Processing Symposium 2020

Abdul Dakkak
Abdul Dakkak

Chen and Dakkak also collaborated on the Best Paper from IEEE IPDPS 2020. This project, titled “XSP: Across-Stack Profiling and Analysis of Machine Learning Models on GPU,” also included Hwu of ECE and Xiong of the IBM T.J. Watson Research Center. Fellow authors included Wei Wei of the Alibaba Group and Lingjie Xu of Biren Research.

Their paper proposes XSP – an “across-stack profiling design that gives a holistic and hierarchical view of Machine Learning (ML) model execution.” The work is derived from the “rapid proliferation of machine learning/deep learning models” that has made “profiling and characterization of ML model performance and increasingly pressing task for both hardware designers and system providers.”

The authors view XSP as a development that “provides insights which would be difficult to discern otherwise.”

Best Demonstration Paper Award, Association for Computational Linguistics 2020

Heng Ji
Heng Ji

Illinois CS PhD student Manling Li was first author of the paper titled “GAIA: A Fine-grained Multimedia Knowledge Extraction System,” which earned ACL 2020 Best Demonstration Paper. Li authored the paper with her Illinois CS advisor, Heng Ji, and fellow authors Alireza Zareian, Ying Lin, Xiaoman Pan, Spencer Whitehead, Brian Chen, Bo Wu, Shih-Fu Chang, Clare Voss, Daniel Napierski and Marjorie Freedman. Other contributors work with institutions like Columbia University, US Army Research Laboratory and the Information Sciences Institute.

Their report presents “the first comprehensive, open source multimedia knowledge extraction system that takes a massive stream of unstructured, heterogeneous multimedia data from various sources as input, and creates a coherent, structured knowledge base.”


Third Best Paper, SIGCSE Technical Symposium 2021

Craig Zilles
Craig Zilles

Illinois CS professor Craig Zilles and PhD student Max Fowler earned third best paper at SIGCSE TS 2021 for the work titled “Superficial Code-guise: Investigating the Impact of Surface Feature Changes on Students' Programming.”

It was noted from SIGCSE TS that the program chairs “selected three best papers … from each of the paper tracks for their accomplishment of high quality, novelty and broad appeal to reviewers.”

Zilles and Fowler worked to better assess student performance on programming questions while also mitigating cheating efforts for computer-based and only assessment efforts. The way to do so, they surmised, is to “produce large pools of questions quickly” for which they “used a permutation strategy to rapidly make new question variants by altering existing questions’ surface features.”

Honorable mention, Best Diversity, Equity & Inclusion Paper Award, American Society for Engineering Education 2020

Lawrence Angrave
Lawrence Angrave

At the American Society for Engineering Education in 2020, Illinois CS professor Lawrence Angrave worked with a team to produce the work “Improving Student Accessibility, Equity, Course Performance, and Lab Skills: How Introduction of ClassTranscribe is Changing Engineering Education at the University of Illinois.”

The authors included fellow Illinois CS professor Hongye Liu, as well as Illinois CS students Zhilin Zhang, Chirantan Mahipal, and Ruihua Sui.

Their paper presented three case studies that “examined the use of ClassTranscribe in a diverse set of undergraduate engineering classes in 2019 and 2020 at the University of Illinois at Urbana-Champaign.” ClassTranscribe is a “video viewing system designed with accessibility and learning in mind.”

IEEE Micro Top Picks 2021

Dimitrios Skarlatos
Dimitrios Skarlatos

Two Illinois CS papers received an award from IEEE Micro Top Picks 2021:

  • First is a paper from Illinois CS PhD graduate Dimitrios Skarlatos, titled “BabelFish: Fusing Address Translations for Containers,” which earned top pick recognition. Skarlatos authored the paper with Illinois CS professor Josep Torrellas, ECE professor Nam Sung Kim, Illinois CS PhD graduate Bhargava Gopireddy, and ECE graduate Umur Darbaz. Their work proposes BabelFish as a “novel architecture to share page translations across containers in the TLB and in page tables.” Their findings reported reduced “latency of containerized data-serving workloads,” and lowered “execution time of containerized compute workloads.”
  • The second paper, titled “Elasic Cuckoo Page Tables: Rethinking Virtual Memory Translation for Parallelism,” earned honorable mention. Skarlatos was again the first author, this time working with Torrellas, Illinois CS professor Tianyin Xu and Illinois CS PhD student Apostolos Kokolis. Their research proved a new method to conduct address translations at a faster rate.

Best Paper Award, IEEE International Symposium on Software Reliability Engineering 2020

AV-Fuzzer: Finding Safety Violations in Autonomous Driving Systems,” earned the Best Paper Award at the IEEE International Symposium on Software Reliability Engineering (ISSRE) 2020. Authors include Guangpeg Li, Yiran Li, Illinois CS PhD student Saurabh Jha, Timothy Tsai, Michael Sullivan, Siva Kumar Sastry Hari (PhD CS ’13), and Illinois ECE professors Zbigniew Kalbarczyk and Ravishankar Iyer.

The work tested production autonomous vehicle technology to find nominal safety bugs.