Lightning Research Introduction Seminars Series 1 (Fall 2015)

14 September 2015 (Monday)

Seminar: Lightning Research Introduction Seminar (1)

Date: 14 September 2015 (Monday)

Time: 4-5pm (light refreshment will be arranged outside the LTF after the seminar)

Venue: LTF (near lifts 25/26), HKUST

Host: Prof. Qiang YANG (Head of Dept.)

Speakers: CSE Faculty Members

1. Fangzhen LIN
2. Qiong LUO
3. Huamin QU
4. Long QUAN
5. Andrew Horner
6. Siu-Wing CHENG
7. Chi-Keung TANG
8. Lei CHEN
9. Qiang YANG

Schedule

Full Video Playback

Abstracts

Talk 1

Speaker: Prof. Fangzhen LIN

Title: "Programs as Agents in First-Order Logic"

Abstract:

I will briefly talk about my current research interests and in particular 
my work on translating programs to first-order logic.

Talk 2

Speaker: Dr. Qiong LUO

Title: "Fast Data Systems on Modern Computers"

Abstract:

Database research has long gone beyond SQL (Structured Query Language) 
engines and branched into various areas of data-intensive algorithms and 
systems. My students and I have been working on how to utilize current 
commodity hardware, such as multicore CPUs, graphics processors (GPUs), and 
solid state drives (SSDs), to speed up data processing and analytical 
tasks. For example, we have developed (1) Mars, a MapReduce System that 
takes advantage of GPUs; (2) Gallop, a GPU-accelerated genome sequence 
analysis system; and (3) Star Flow, a parallel processing pipeline for 
astronomcial observation data.

Talk 3

Speaker: Prof. Huamin QU

Title: "Making Sense of Big Data with Visual Analytics"

Abstract:

In this talk, I will briefly introduce the visualization research 
conducted in my group, with focuses on urban informatics, social network 
analysis, e-learning, and text visualization. For more information, please 
check http://vis.cse.ust.hk/

Talk 5

Speaker: Prof. Andrew HORNER

Title: "Emotional Characteristics of Musical Instruments Sounds"

Abstract:

The main goal of our recent research is to determine the relationship 
between music emotion and sound color in music and musical instruments. In 
particular, we are determining how instruments vary in emotional 
characteristics due to pitch, dynamics, and sound color. We use listening 
tests, sound analysis, and statistical analysis to determine significant 
differences in emotional characteristics over different pitches, dynamics, 
and instruments. For example, our current research on the piano indicates 
that the high-soft notes of the piano are most romantic, the low-loud 
notes most heroic, and the lowest- and highest- loud notes most scary. 
This work has many real-world applications since emotional characteristics 
can be manipulated in a recording, performance, or composition by 
emphasizing instruments, pitches, and dynamics that are comparatively 
stronger in representing these characteristics.

Talk 6

Speaker: Prof. Siu-Wing CHENG

Title: "My Work/Computational Geometry"

Abstract:

I will give a short overview of my recent research in computational 
geometry.

Talk 7

Speaker: Prof. Chi-Keung TANG

Title: "Efficient Square Localization for Accurate Object Detection"

Abstract:

The key contribution is the compact square object localization, which
relaxes the sliding window from testing windows of different combinations
of aspect ratios.

Talk 8

Speaker: Dr. Lei CHEN

Title: "Spatial Crowd Sourcing, Opportunities and Challenges"

Abstract:

Crowdsourcing is a new computing paradigm where humans are enrolled 
actively to participate into the procedure of computing, especially for 
the tasks that are intrinsically easier for human than for computers. Not 
surprisingly, with the development of mobile Internet, the magic power of 
crowdsourcing is now expanding to physical world, where each user is 
treated as a mobile computing unit that can be activated and guided for 
certain tasks. Such practice is in general termed as spatial 
crowdsourcing, featuring task dispatching and dynamic pricing as its core 
technical niches. Therefore, it serves as the fundamental prototype of a 
cluster of industrial applications like Citizen Sensing(Waze), P2P ride- 
sharing(Uber), Real-time O2O service(Instacart, Postmates) and so on.

In this talk, I will first show a short video of spatial crowdsourcing. 
Then, I will introduce the theoretical and practical development our 
spatial crowdsourcing project, G-mission. Finally, I will talk about some 
interesting future works on G- mission.

Talk 9

Speaker: Prof. Qiang YANG

Title: "Transfer Learning on Big Data"

Abstract:

We are doing research on making transfer learning systems practical on big 
data. Our aim is to integrate the data from diverse problem domains and to 
facilitate feature engineering to enable effective model building.  In 
this talk, I will highlight some of our works in heterogeneous transfer 
learning, transitive transfer learning and recommendation systems.