Select a calendar:
Filter October Events by Event Type:
SUNMONTUEWEDTHUFRISAT
University Calendar
Events for October
-
Towards Mobile Emoji Prediction from Speech and Textual Captions
Tue, Oct 03, 2017 @ 02:30 PM - 05:00 PM
Ming Hsieh Department of Electrical and Computer Engineering
University Calendar
Bio:
Luis Marujo is a Research Scientist at Snap Inc. Prior to joining the Snap Research team in 2016, he completed his dual-degree Ph.D. in Language Technologies ('15) from Carnegie Mellon University (CMU) and the Instituto Superior Técnico (IST), Portugal. He also obtained a MSc. ('12) in Language Technologies from CMU. He holds MSc. ('09) and BSc. ('07) in Computer Science and Engineering from IST. He was awarded the best poster award at the S3MR 2011.
Abstract:
Emojis are very popular ideograms used to either concisely communicate or complement the information of a text with an emotion or visual concept. Emojis are mainly use on mobile devices due to the availability of intuitive emoji keyboards. They are very popular in social media, but they have not been explored from a Speech Processing point of view. In this work, we investigate mobile-friendly multimodal approaches that use audio, speech, and textual captions to predict emojis in public video Snaps. The key idea of pipeline is to translate the input signals into text and keywords in order to conduct the analysis in a textual space. Our emoji prediction pipeline includes speech transcription, keyword spotting and dictionary based classification.
In addition we use music detection and language identification to filter the content. Our experimental results indicate that our approach using speech transcription or a list of pre-selected words from keyword spotting provide comparable information to what is found in textual captions for emoji prediction. This is an important result as it allows us to suggest emojis to snap videos before the user starts typing a caption. Our initial results also indicate that combining both textual captions with speech output can improve emoji recommendation results beyond using only textual captions.Location: Hughes Aircraft Electrical Engineering Center (EEB) - 132
Audiences: Everyone Is Invited
Contact: Benjamin Paul
-
STEM Spotlight on the Center for Advanced Manufacturing
Fri, Oct 06, 2017 @ 10:00 AM - 01:00 PM
USC Viterbi School of Engineering, Viterbi School of Engineering K-12 STEM Center
University Calendar
As part of the STEM Spotlight on the Department of Aerospace and Mechanical Engineering, the Center for Advanced Manufacturing will open its labs to high school visitors on October 6, on National Manufacturing Day. The STEM Spotlight on AME continues on the University Park campus research labs on October 12. https://viterbipk12.usc.edu/cam-spotlight/ Presented by Viterbi Adopt-a-School, Adopt-a-Teacher (VAST) preK-12 STEM Education Outreach.
Location: 2727 S Flower St
Audiences: K-12 Schools pre-registered
Contact: Katie Mills
-
STEM Spotlight on Aerospace & Mechanical Engineering
Thu, Oct 12, 2017 @ 09:00 AM - 01:30 PM
USC Viterbi School of Engineering, Viterbi School of Engineering K-12 STEM Center
University Calendar
High school students will tour the research labs, student design studios, and demo spaces of the Department of Aerospace and Mechanical Engineering as part of the STEM Spotlight series presented by USC Viterbi Adopt-a-School, Adopt-a-Teacher (VAST). On 10/6, the Center for Advanced Manufacturing will also provide tours for visiting students. For more info, see https://viterbipk12.usc.edu/research/stem-spotlight/aerospace-mechanical-engineering/
Location: Robert Glen Rapp Engineering Research Building (RRB) - Several locations: RRB, RTH, OHE, SHL, BHE
Audiences: K-12 Schools pre-registered
Contact: Katie Mills
-
PhD Defense- You Kyu Lee
Thu, Oct 19, 2017 @ 01:00 PM - 03:00 PM
Thomas Lord Department of Computer Science
University Calendar
Date: Thu, Oct 19, 2017 @ 1:00 PM - 3:00 PM
Location: Hedco Chemical Engineering (HED) - 103
PhD Candidate: Youn Kyu Lee
Committee: Nenad Medvidovic (CS, chair), William G.J. Halfond (CS), Sandeep Gupta (EE)
Title: Reducing Inter-Component Communication Vulnerabilities in Event-Based Systems
Abstract:
Event-based system (EBS) has become popular because of its high flexibility, scalability, and adaptability. These advantages are facilitated by its reliance on implicit invocation and implicit concurrency. Specifically, in EBS, components may not know the consumers of the events they publish, nor do they necessarily know the producers of events they consume. This communication mechanism is based on non-determinism in event processing, which can introduce inherent security vulnerabilities into a system referred to as event attacks. Event attack is a particular type of attack that can abuse, incapacitate, and damage a target system by exploiting the system's event-based communication model. Different types of event attacks have been identified in a range of domains to date. It is hard to prevent event attacks because they are administered in a way that does not differ from ordinary event-based communication in general. While a number of techniques have focused on security threats in EBS, they do not appropriately resolve the event attack problems or suffer from inaccuracy in detecting and preventing event attacks. Furthermore, fundamental security flaws, which can be exploited by event attacks, have not been clearly identified yet. In order to address the risk of event attacks, this dissertation presents four main approaches: (1) a new taxonomy for security flaws in EBS, which can serve as a basis for resolving event attack problems; (2) SEALANT (Security for End-users of Android via Light-weight ANalysis Techniques), a novel protection mechanism for Android, one of the most widely used event-based platforms; (3) SCUTUM (SeCUrity for evenT-based systems implemented Using MOM platforms), a novel vulnerability detection technique for EBSs that are implemented by using message-oriented middleware platforms; and (4) ViVA (Visualizer for eVent-based Architectures), a new visualization technique for monitoring and identifying security vulnerabilities in EBS.
Location: 103
Audiences: Everyone Is Invited
Contact: Lizsl De Leon
