Home

Call for Papers

Call for Workshops

Call for Challenges

Call for Tutorials

Important Dates

Committees

Reviewers

Conference

Program

Workshops

Challenges

Tutorials

Keynote Speakers

Registration

Venue

Travel

Tour/Attractions

Accommodation

Sponsorship

Past AVSS

Authors

Submission

Accepted Papers

Camera-Ready

Presentation and Poster Instructions

Program

We encourage paper submissions and welcome you to attend AVSS 2019. Please see Call for Papers for details.

Program Overview:

  9/18 Wed 9/19 Thr 9/20 Fri 9/21 Sat
Morning Keynote I Keynote III Keynote V

Workshops
Challengs

Tutorials

Oral Session I Oral Session III Oral Session V
Poster Session I Poster Session III Poster Session V
Lunch        
Afternoon Keynote II Keynote IV City Tour

Workshops
Challengs

Tutorials

Oral Session II Oral Session IV
Panel I Panel IV
Poster Session II Oral Session IV
Dinner Reception Banquet   Adjoint

Note that ICIP 2019 Taipei is a back-to-back conference on 9/22 (Sun) to 9/25 (Wed).

Program Schedule

All Conference and Workshop meeting rooms are in the Gong-Cheng Building (公誠樓) of University of Taipei (UTaipei) - Bo-Ai Campus (臺北市立大學博愛校區). See map in the Venue page.

Main Conference Day 1 (9/18 Wednesday) at Conference Room 3 (會議室三)

8:20 to 8:40 Opening Ceremony
8:40 to 9:40 Keynote I - Prof. Mubarak Shah
9:40 to 10:00 Coffee Break
10:00 to 11:00 Oral I (3 papers, session chair: Prof. Mubarak Shah)
11:00 to 12:00 Poster I (8+3 posters)
12:00 to 1:30 Lunch
1:30 to 2:30 Keynote II - Prof. Jenq-Neng Hwang
2:30 to 3:30 Oral II (3 papers, session chair: Prof. Jenq-Neng Hwang)
3:30 to 3:50 Coffee Break
3:50 to 4:30 Panel Discussion I (Industrial Perspectives on the 3rd wave of AI)
4:30 to 5:30 Poster II (8+3 posters)
5:30 to 8:00 Reception at the Venue

Main Conference Day 2 (9/19 Thursday) at Conference Room 3 (會議室三)

8:20 to 8:40 Opening (Announcements)
8:40 to 9:40 Keynote III - Dr. Jeff Alstott
9:40 to 10:00 Coffee Break
10:00 to 11:00 Oral III (3 papers, chair: Dr. Jeff Alstott)
11:00 to 12:00 Poster III (8+3 posters)
12:00 to 1:30 Lunch
1:30 to 2:30 Keynote IV - Dr. Hong-Yuan Mark Liao
2:30 to 3:30 Oral IV (3 papers, session chair: Dr. Hong-Yuan Mark Liao)
3:30 to 3:50 Coffee Break
3:50 to 4:30 Panel Discussion II (Ethics & Forensics of AI)
4:30 to 5:30 Poster IV (8+3 posters)
5:30 to 7:30 Coffee Break and transport to Banquet Restaurant
7:30 to 10:30 Banquet

Main Conference Day 3 (9/20 Friday) at Conference Room 3 (會議室三)

8:20 to 8:40 Opening (Announcements)
8:40 to 9:40 Keynote V - Prof. Rama Chellappa
9:40 to 10:00 Coffee Break
10:00 to 11:00 Oral V (3 papers, session chair: Prof. Rama Chellappa)
11:00 to 12:00 Poster V (8+3 posters)

12:00 to 5:00 City Tour

Please sign up during Registration or in the conference no later than the Banquet night, so we can better control head counts for the organization of these tours. We encourage you to bring friends and family members to join the half-day trip. Note that we will provide student volunteering tour guides, but all the expense of the tour (transportation, food, ticket) is is out of your own pocket.

Workshops and Tutorials (9/21 Saturday)

Refer to Workshops and Tutorials for details. Note that each workshop might have difference start and end time, please refer to each workshop website for details. We maintain the same Coffee Break and Lunch time as the main conference across all workshops and tutorials:

9:40 to 10:00 Coffee Break
12:00 to 1:30 Lunch
3:30 to 3:50 Coffee Break

Tutorial I: Automatic Detection of Abnormal Events in Surveillance Videos fro Online Processing and Low-cost Devices

Tutorial II: Deep Learning for Video Compression and Understanding

Workshop (morning): Biometrics for Smart Cities.

Workshop (morning): The 3th International Workshop on Traffic and Street Surveillance for Safety and Security (IWT4S)

Workshop (morning): The 2nd International Workshop on Small-Drone Surveillance, Detection and Counteraction Techniques (WOSDETC)

Workshop (afternoon): The Content-Aware Video Analysis (CAVA)Workshop

Workshop (afternoon): Visual Computing for Computer Vision and Intelligent Transportation System (VCAVSS)

Workshop (afternoon): The Deep Learning in Activity Monitoring (DLAM) Workshop

Panel Discussions

Panel Discussion I: Industrial Perspectives on the 3rd wave of AI

Or actually the discussion is more about Philosophy vs. the 3rd Wave of AI. Refer to Peter Tu's inspirational slides at 2019_AVSS_Third_Wave_AI _Panel_Discussion.pdf.

9/18 Wednesday 3:50 to 4:30pm at Conference Room 3 (會議室三)

Artificial Intelligence has been characterized in terms of waves. During the first wave we explored the idea of logical reasoning. Success in the second wave is primarily based on statistical inference. Those that are calling for the third wave of AI argue for systems that when confronted by unforeseen circumstances can reason over contextual cues and then do the right thing. The questions that we will consider during this panel discussion will revolve around Industries response to the third wave of AI. Topics of discussion may include: the grounding problem, associative and episodic memory, narratives, causality, analogies and questions regarding intentionality.

Panelists (including but not limited to):

Panel Discussion II: Ethics & Forensics of AI

See kick off slides at AVSS10_Panel2_Ethics_Forensics.pdf.

9/19 Thursday 3:50 to 4:30pm at Conference Room 3 (會議室三)

The development of AI technology, in particularly, deep learning and their applications in computer vision, has brought forth significant advances to surveillance. With the improvement in performance, they also introduce new challenges to the research community. In particular, there are three areas that are of important concerns:

  1. biases: the heavy reliance of the deep learning based systems on annotated training data also introduce biases to under-represented social groups in the trained system, thus their fairness with regards to established social values is an important concern recently.

  2. malicious attacks: deep learning based systems are fragile and susceptible to malicious attacks in the form of training data poisoning, deep network backdoor or adversarial examples. As such the security of the deep learning based surveillance systems needs to be addressed.

  3. manipulations: deep learning based methods can also be used to create or tamper digital audiovisual signals (e.g., DeepFakes) misleading surveillance systems and we need to develop effective forensic technology to counter the fake media.

Furthermore, the increasing scales of automatic video surveillance, especially face recognition has also raised significant concerns about the ethical issues in using such technologies. The technical community should also pay attention to the general responsibility to the society.

This panel will provide the audience a comprehensive overview of these issues and focus on an in-depth discussion about their impact to future research directions.

Panelists (including but not limited to):

 

Full Program and Presentation Schedule

Note that all oral papers will also present their posters in their respective session.

Day 1 (9/18 Wed.) Session I (Oral I + Poster I)

Day 1 (9/18 Wed.) Session II (Oral II + Poster II)

Day 2 (9/19 Thr.) Session III (Oral III + Poster III)

Day 2 (9/19 Thr.) Session IV (Oral IV + Poster IV)

Day 3 (9/20 Fri.) Session V (Oral V + Poster V)