Category Archives: Uncategorized

Li in CICS Thurs. 10/8 12:00-1:00

Yunzhu Li will give a talk at Machine Learning and Friends Lunch talk series.

Bio: “I am a third-year Ph.D. student at Computer Science and Artificial Intelligence Laboratory (CSAIL) at  MIT, advised by Prof. Antonio Torralba and Prof. Russ Tedrake. Before coming to MIT, I received the B.S. in Computer Science from Peking University in 2017.

My research interests lie in computer vision, machine learning, and robotics. In particular, I am interested in how we can enable better robotic manipulation skills via learning-based dynamics modeling and multimodal perception.”

About

The Machine Learning and Friends Lunch (MLFL) series is sponsored by our friends at Oracle Labs.

MLFL is a lively and interactive forum held weekly where friends of the UMass Amherst machine learning community can gather virtually and give or hear a 50-minute presentation on recent machine learning research.

What is it?   A gathering of students/faculty/staff with broad interest in the methods and applications of machine learning.
When is it?    Thursdays 12:00pm to 1:00pm, via Zoom
Who is invited?   Everyone is welcome.
More info? Email cds-info@cs.umass.edu with questions or suggestions.

Harwath in CICS Thurs. 10/1 12:00-1:00pm

Thursday, October 1, 2020 – 12:00 David Harwath (University of Texas at Austin): Multimodal Perception

My research interests are in the area of machine learning for speech and language processing. The ultimate goal of my work is to discover the algorithmic mechanisms that would enable computers to learn and use spoken language the way that humans do. My approach emphasizes the multimodal and grounded nature of human language, and thus has a strong connection to other machine learning disciplines such as computer vision. 

While modern machine learning techniques such as deep learning have made impressive progress across a variety of domains, it is doubtful that existing methods can fully capture the phenomenon of language. State-of-the-art deep learning models for tasks such as speech recognition are extremely data hungry, requiring many thousands of hours of speech recordings that have been painstakingly transcribed by humans. Even then, they are highly brittle when used outside of their training domain, breaking down when confronted with new vocabulary, accents, or environmental noise. Because of its reliance on massive training datasets, the technology we do have is completely out of reach for all but several dozen of the 7,000 human languages spoken worldwide.

In contrast, human toddlers are able to grasp the meaning of new word forms from only a few spoken examples, and learn to carry a meaningful conversation long before they are able to read and write. There are critical aspects of language that are currently missing from our machine learning models. Human language is inherently multimodal; it is grounded in embodied experience; it holistically integrates information from all of our sensory organs into our rational capacity; and it is acquired via immersion and interaction, without the kind of heavy-handed supervision relied upon by most machine learning models. My research agenda revolves around finding ways to bring these aspects into the fold.

Prior to joining UT, I worked as a research scientist at MIT CSAIL from 2018 to 2020. I recieved my PhD in 2018 from the Spoken Language Systems Group at MIT CSAIL, under the supervision of Jim Glass.

About

The Machine Learning and Friends Lunch (MLFL) series is sponsored by our friends at Oracle Labs.

MLFL is a lively and interactive forum held weekly where friends of the UMass Amherst machine learning community can gather virtually and give or hear a 50-minute presentation on recent machine learning research.

What is it?   A gathering of students/faculty/staff with broad interest in the methods and applications of machine learning.
When is it?    Thursdays 12:00pm to 1:00pm, via Zoom
Who is invited?   Everyone is welcome.
More info? Email cds-info@cs.umass.edu with questions or suggestions.

Pérez-Leroux and Roberge in LARC Fri 10/2 at 9am

Please join us this Friday at LARC for a presentation by Ana Pérez-Leroux and Yves Roberge (Univ. of Toronto):

” Single form, multiple functions: Evidence of L1 acquisition of  Japanese recursive NO”

The meeting starts at 9:00 (Estern Time – US).

Join Zoom Meeting https://umass-amherst.zoom.us/j/92789104594?pwd=bFQzRTZlYXJVaHh5MG0wcElzOXBzUT09 Meeting ID: 927 8910 4594 Passcode: research

Sensus: Constructing Meaning in Romance 9/26-27

We are pleased to announce that the first meeting of Sensus, a workshop on the formal semantics and pragmatics of Romance languages, will take place virtually at the University of Massachusetts Amherst, on September 26-27 2020.

Our intention with this workshop is to bring together scholars in the growing community dedicated to the study of the construction of meaning within Romance languages. We hope this will become a regular venue for researchers with common interests in Romance semantics and pragmatics.

Invited speakers: Luis Alonso-Ovalle (McGill University), Mariapaola D’Imperio (Rutgers University), Donka Farkas (University of California, Santa Cruz).

Registration is free. Please register to obtain Zoom link shortly before conference.

– On behalf of your Sensus organizers: Ana Arregui, María Biezma, Vincent Homer and Deniz Özyıldız)

Wieting in CICS Thurs. 9/24 12:00-1:00pm

John Wieting will give a talk in Machine Learning and Friends Lunch (MLFL) series.

What is it?   A gathering of students/faculty/staff with broad interest in the methods and applications of machine learning.
When is it?    Thursdays 12:00pm to 1:00pm, via Zoom
Who is invited?   Everyone is welcome.
More info? Email cds-info@cs.umass.edu with questions or suggestions.

John is a Ph.D. student at Carnegie Mellon University supervised by Taylor Berg-Kirkpatrick and Graham Neubig, studying machine learning, learning theory, optimization, natural language processing, and computer vision. Currently, John’s research has focused on machine learning and natural language processing.

The Machine Learning and Friends Lunch (MLFL) series is sponsored by our friends at Oracle Labs.

MLFL is a lively and interactive forum held weekly where friends of the UMass Amherst machine learning community can gather virtually and give or hear a 50-minute presentation on recent machine learning research.

Experiments in linguistic meaning 9/16-9/18

The inaugural conference on Experiments in Linguistic Meaning (ELM*) is being hosted virtually by the University of Pennsylvania, September 16-18, 2020.

More information can be found on their website: https://www.elm-conference.net/#:~:text=We%20are%20excited%20to%20announce,September%2016%2D18%2C%202020.&text=Also%2C%20be%20assured%20that%20we,to%20welcoming%20you%20to%20ELM!

Vasishth in linguistics colloquium Friday 9/25 at 3:30

Shravan Vasishth (Potsdom) will give a colloquium talk in linguistics on 9/25 at 3:30. Below is the abstract and zoom information.

———— Abstract————
Twenty years of retrieval models
Shravan Vasishth (vasishth.github.io)

After Newell wrote his 1973 article, “You can’t play twenty questions
with nature and win”, several important cognitive architectures
emerged for modeling human cognitive processes across a wide range of
phenomena. One of these, ACT-R, has played an important role in the
study of memory processes in sentence processing. In this talk, I
will talk about some important lessons I have learnt over the last 20
years while trying to evaluate ACT-R based computational models of
sentence comprehension. In this connection, I will present some new
results from a recent set of sentence processing studies on Eastern
Armenian.

Reference:
Shravan Vasishth and Felix Engelmann. Sentence comprehension as a
cognitive process: A computational approach. 2021. Cambridge
University Press.
https://vasishth.github.io/RetrievalModels/

———-Zoom Invitation—————
Sep 25, 2020 03:30 PM Eastern Time (US and Canada)

Register in advance for this meeting:
https://umass-amherst.zoom.us/meeting/register/tJUldemurz4oGdAo6hV69nh4k3y82zRiLVZB

White in Cognitive Brown bag Thurs 9/9 at noon

Alex White will give a talk on visual word recognition in Cognitive Brown bag over zoom on Thursday, 9/9 at noon. Below is the abstract and zoom information. Everyone is welcome!

A bottleneck for visual word recognition in brain and behavior

Reading is a demanding task, constrained by inherent processing capacity limits. Do those capacity limits allow for multiple words to be recognized in parallel? I will first present a series of behavioral experiments that measured accuracy for recognizing briefly flashed and masked words. The question is: when given the amount of time required to recognize one word, can people divide attention and recognize two words simultaneously?  We consistently found that accuracy was so impaired that it supported a serial processing model: on each trial, subjects could recognize one word but had to guess about the other. We concluded that a fundamental bottleneck allows only one word to be fully processed at a time. I will then present a neuroimaging study designed to identify the source of that bottleneck, focusing on two sub-regions of the “visual word form area,” a critical component of the brain’s reading circuitry. The data suggest that the posterior sub-region can encode two words in parallel spatial channels. In contrast, activity in the anterior portion of the visual word form area is consistent with a bottleneck. Therefore, the visual system can process two words in parallel up to a relatively late stage in the recognition hierarchy. I look forward to discussion with the audience about what implications these results have for theories of reading.

Join Zoom Meeting
https://umass-amherst.zoom.us/j/97489074821?pwd=MG9Nbk9OS1g4N3RXMVZab3duQUtSQT09

Meeting ID: 974 8907 4821
Passcode: PBScog