1st Workshop on Computer Vision for Global Challenges @ CVPR 2019
June 16th 2019, Grand ballroom A, Long Beach Convention Center


The first Workshop on Computer Vision for Global Challenges will be held on June 16th, 2019, in conjunction with the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), in Long Beach, California. It will be a full-day event and will feature invited speakers, poster and spotlight presentations, a panel discussion and a mentoring/networking dinner.

We have selected 17 challenge winners to present their works, interact, inspire and get inspired by the computer vision community. We hope these challenges are not just applications that may affect billions, but also contribute to computer vision research by unveiling current limitations and biases. The selected challenge proposals come from 15 different countries and were selected among 100+ applications. From crop diseases to air pollution, from natural disaster prevention to wildlife poaching prevention, from dyslexia to malaria diagnosis, the challenges cover diverse areas like agriculture, health, education and meteorology, to name a few.

We have further invited a cohort of more than 15 International Development experts, making CV4GC the first workshop that tries to bridge the gap between the computer vision research, and international development communities.


Distribution of accepted challenge applications.




Workshop Program
June 16th 2019, Grand ballroom A, Long Beach Convention Center

Links to all presented material will be added soon. Click on the button below to toggle session details.

Start

End

Session

8:45 AM

9:00 AM

Welcome remarks

9:00 AM

9:30 AM

Keynote Presentation

- Collaboration and Co-Creation in Applying Computer Vision for Social Impact
Padmanabhan Anandan, CEO, Wadhwani AI

9:30 AM

10:00 AM

Invited Talks and Discussion - 1

- Missing Pixels: The Need for Inclusion in AI
Deborah Raji, Google AI

- Using satellite imagery to analyze spatial apartheid in South Africa
Raesetje Sefala, Wits University

10:00 AM

10:30 AM

Lightning talks from International Development Experts


- Creating labeled training datasets as a way to encourage native CV applications in emerging markets.
Sue Marques, Rockefeller Foundation

- Connected and digital health in a global perspective
Ben Pierson and Arunan Skandarajah, Bill and Melinda Gates Foundation

- Fighting counterfeiting of life-impacting products like medicines and agro-inputs, and securing cold-chains with CV in Africa, South Asia & Middle East
Bright Simons, Mpedigree

- Predicting deforestation and detecting illegal logging with CV in Chile
Micah Melnyk, World Bank/Singularity

- Disruptive tech for financial inclusion - or how to help East African shopkeepers manage their business?
Matt Grasser, Bankable Frontiers Associates

- Project Connect - how to map every school in the world without data
Zhuang-Fang Yi, Development Seed

10:30 AM

10:45 AM

Coffee break

10:45 AM

11:15 AM

Invited Talks and Discussion - 2

- The role of context in the context of computer vision for development challenges
Ernest Mwebaze, Google AI Accra

- Misinformation: a looming threat to sustainable development in Latin America.
Claudia Flores Saviaga, West Virginia University

11:15 AM

12:00 PM

Selected Oral Presentations

- Creating xBD: A Dataset for Assessing Building Damage from Satellite Imagery
Ritwik Gupta, Bryce Goodman, Nirav Patel, Ricky Hosfelt, Sandra Sajeev, Eric Heim, Jigar Doshi, Keane Lucas, Howie Choset, Matthew Gaston

- Deep Landscape Features for Improving Vector-borne Disease Prediction

Nabeel Abdur Rehman, Umar Saif, Rumi Chunara

- Does Object Recognition Work for Everyone?

Terrance de Vries, Ishan Misra, Changhan Wang, Laurens van der Maaten

- Infant-Prints: Fingerprints for Reducing Infant Mortality

Joshua Engelsma, Debayan Deb, Anil Jain, Anjoo Bhatnagar, Prem Sewak Sudhish

12:00 PM

1:00 PM

Lunch

1:00 PM

1:30 PM

Invited Talks and Discussion - 3

- The use of computer vision in understanding and measuring human well-being
Marshall Burke (Stanford University)

- Title: TBD
Jitendra Malik (UC Berkeley and Facebook AI)

1:30 PM

2:45 PM

Global Challenges - Lightning talks

Health
- A rapid malaria diagnosis test using a mobile phone
Martha Shaka (The University of Dodoma, Tanzania)
- Early Detection and Diagnosis of Breast Cancer using Image Processing and Machine Learning
Maleika Heenaye-Mamode Khan (University of Mauritius, Mauritius)
- Towards Detecting Dyslexia in Handwriting Using Neural Networks
Katie Spoon (Indiana University Bloomington, USA)
- "Bytes against bites": First AI-based app for snakebite management
Rafael Ruiz de Castañeda (Institute of Global Health, Faculty of Medicine, University of Geneva, Switzerland)
- Computer Vision for Childbirth Progression Monitoring: Cervical Dilation Assessment
Temitope Oluwatosin Takpor (Covenant University, Nigeria)

Agriculture
- Smartphone-based application for the identification of diseases in crops using deep convolutional neural networks: AgroTIC + SMART
Ariolfo Camacho Velasco (Universidad Industrial de Santander, Colombia)
- CareGreenUnit and CareGreenMonitering system: AI Support Decision Making System in Smart Agriculture
Azeddine EL HASSOUNY (Mohammed V University in Rabat, Morocco)
- Intelligent Crop Health and Pest Detection
Mohsen Ali (Information Technology University, Pakistan)
- A Computer Vision Tomato Pest Assessment and Prediction Tool
Denis Pastory Rubanga (Tokyo University of Agriculture, Japan)
- Inferring Crop Pests and Diseases from Image Soil Data and Soil Properties
Claire Babirye (Uganda Technology and Management University, Uganda)

Remote Sensing
- Eliminating Bonded Labour in Brick Kilns of South Asia
Murtaza Taj (Lahore University of Management Sciences, Pakistan)
- Automatic Illegal mining detection using satellite imagery and computer vision techniques
Leticia Leonor Pinto Alva (Universidad Catolica San Pablo, Peru)
- Poverty Prediction Using Satellite Imagery
Chaiyaphum, Siripanpornchana (NECTEC Thailand, Thailand)
- Computer Vision for Drought Resilience
Andrew Hobbs (UC Davis, USA)
- Storm Nowcasting Improvement by Machine Learning
Suzanna Maria Bonnet de Oliveira Martins (Federal University of Rio de Janeiro, Brazil)

Wildlife & Air Quality
- Automated Wildlife Monitoring, Tracking and Poacher Detection
Wilhelmina Ndapewa Nekoto (Dundee Precious Metals, Namibia)
- Nowcasting Air Quality using Social Media Imagery
Anthony Mockler (UN Global Pulse Lab Jakarta, Indonesia)

2:45 PM

3:30 PM

Global Challenges - Breakout Session

3:30 PM

4:15 PM

Posters and Coffee

- Building High Resolution Maps for Humanitarian Aid and Development with Weakly- and Semi-Supervised Learning
Derrick Bonafilia (Facebook); James Gill (Facebook); Saikat Basu (Facebook); David Yang (Facebook)
- Creating xBD: A Dataset for Assessing Building Damage from Satellite Imagery

Ritwik Gupta (Carnegie Mellon University Software Engineering Institute); Bryce Goodman (Defense Innovation Unit); Nirav Patel (Defense Innovation Unit); Ricky Hosfelt (Carnegie Mellon University Software Engineering Institute); Sandra Sajeev (Carnegie Mellon University Software Engineering Institute); Eric Heim (Carnegie Mellon University Software Engineering Institute); Jigar Doshi (CrowdAI, Inc.); Keane Lucas (Joint Artificial Intelligence Center); Howie Choset (Carnegie Mellon University); Matthew Gaston (Carnegie Mellon University Software Engineering Institute)
- Weakly Labeling the Antarctic: The Penguin Colony Case

Hieu Le (Stony Brook University); Bento Goncalves (Stony Brook University); Dimitris Samaras (Stony Brook University); Heather Lynch (Stony Brook University)
- Towards Autonomous Mining via Intelligent Excavators

Hooman Shariati (Motion Metrics International); Anuar Yeraliyev (Motion Metrics International); Burhan Terai (Motion Metrics International); Shahram Tafazoli (Motion Metrics International); Mahdi Ramezani (Motion Metrics International)
- DisplaceNet: Recognising Displaced People from Images by Exploiting Dominance Level

Grigorios Kalliatakis (University of Essex, UK); Shoaib Ehsan (University of Essex); Maria Fasli (University of Essex, UK); Klaus D McDonald-Maier (University of Essex)
- Tiny-Inception-ResNet-v2: Using Deep Learning for Eliminating Bonded Labors of Brick Kilns in South Asia

Usman Nazir (LUMS); Numan Khurshid (LUMS); Muhammad Bhimra (LUMS); Murtaza Taj (LUMS)
- Deep Landscape Features for Improving Vector-borne Disease Prediction
Nabeel Abdur Rehman (New York University); Umar Saif (United Nations - Pakistan); Rumi Chunara (New York University)
- See the E-Waste! Training Visual Intelligence to See Dense Circuit Boards for Recycling
Ali Jahanian (MIT); Quang Le (MIT); Kamal Youcef-Toumi (MIT); Dzmitry Tsetserukou (skoltech)
- Does Object Recognition Work for Everyone?
Terrance de Vries (University of Guelph); Ishan Misra (Facebook AI Research); Changhan Wang (Facebook AI Research); Laurens van der Maaten (Facebook)
- Towards equitable access to information and opportunity for all: mapping schools with high-resolution Satellite Imagery and Machine Learning

Zhuangfang Yi (Development Seed); Naroa Zurutuza (UNICEF); Drew Bollinger (Development Seed); Manuel Garcia-Herranz (UNICEF); Dohyung Kim (UNICEF)
- Infant-Prints: Fingerprints for Reducing Infant Mortality
Joshua Engelsma (Michigan State University); Debayan Deb (Michigan State University); Anil Jain (Michigan State University); Anjoo Bhatnagar (Saran Ashram Hospital); Prem Sewak Sudhish (Dayalbagh Engineering Institute)
- Semantic Segmentation of Crop Type in Africa: A Novel Dataset and Analysis of Deep Learning Methods
Rose Rustowicz (Stanford University); Robin Cheong (Stanford); Lijing Wang (Stanford University); Stefano Ermon (Stanford University); Marshall Burke (Stanford University); David Lobell (Stanford University)
- Detecting Roads from Satellite Imagery in the Developing World

Yoni Nachmany (Radiant Earth Foundation); Hamed Alemohammad (Radiant Earth Foundation)
- Predicting City Poverty Using Satellite Imagery
Simone Piaggesi (ISI Foundation / Bologna University); Laetitia Gauvin (ISI); Michele Tizzoni (ISI Foundation); Ciro Cattuto (ISI Foundation); Natalia Adler (UNICEF); Stefaan Verhulst (The GovLab); Andrew Young (The GovLab); Rhiannan Price (DigitalGlobe); Leo Ferres (UDD); Andrè Panisson (ISI Foundation)

4:15 PM

5:00 PM

Panel Discussion

How Can These New Challenges Benefit Computer Vision?

- Jitendra Malik (UC Berkeley and Facebook AI)
- Timnit Gebru (Google AI)
- Peter Eckersley (Partnership on AI)
- Ernest Mwebaze (Google AI Accra)
- Laura Sevilla-Lara (University of Edinburgh)

5:00 PM

5:10 PM

Closing remarks

7:00 PM

10:30 PM

Dinner @ L’ Opera (by invitation only)




Motivation

Computer vision technology, a sub-class of Artificial Intelligence technologies, has recently made rapid progress in several subareas (e.g., object segmentation, video classification), achieving a level of performance that was unexpected just a few years ago. This technology has opened possibilities in many real world domains including transportation (self-driving cars, drones), entertainment (shopping, phones), safety (security systems), which are used on a daily basis to make our lives more efficient and safer. While these applications give meaning and value to our technology, their focus and presence in certain geographical regions or markets injects biases in the datasets, tasks, and ultimately direction of the advancement of the field.

However, increased global expansion of Internet and mobile devices (network connected computing devices) offers a unique opportunity for the computer vision community to both expand target markets and geographically diverse applications and input and training biases.

Thus, widening the scope of computer vision applications to address global problems could lead to a win for both the computer vision community and the organizations and individuals working to address global challenges: vision could positively impact the lives of 6 billion people in the Global South by developing applications for remote extension services for farming communities, digital health care delivery, or disaster readiness and mapping of informal settlements for example - and these populations could help reveal some of the blind spots and biases in the current computer vision datasets, tasks, and practices.

Why has this opportunity not been seized before? We argue that one of the main obstacles is the disconnection between domain experts: those who are close to the problems on the ground, and those who have knowledge about technical solutions. This disconnect might be driven by geographical divide, differences in language and taxonomy, or might come from the lack of a accessible forum to find each other. We propose this initiative as a first step to bridge the gap between these two communities. In particular our goals are:




Call for Challenges

Do you have an idea for a computer vision task that would impact the lives of many? Have you identified the limitations of a vision technique because of the geographical bias of the data you are using it on? Is there an application of computer vision that would be helpful to your community? Are you looking for potential vision expert partners or feedback on your idea? Apply for the Call of Challenges, and come and experience the premier computer vision conference, and participate in an active discussion with the top vision researchers. Selected proposals will be presented as spotlight and/or posters during the workshop.

Recipients of accepted proposals may receive full travel grants (transportation, accommodation, registration and meals) to attend the CVPR conference, tutorials and workshops. The review process is single-blind and immigration letters will be provided.
Researchers or practitioners based in developing regions are strongly encouraged to apply.

Travel grants and US Visa: Keeping in mind how time-consuming the Visa processes can be, we expect to announce decisions for applicants that do not have a US Visa around mid-April. This will hopefully allow enough time for the visa process to finish on time.

The submission deadline for the Call for Challenges has now passed (was March 25th, 2019). We received 105 applications in total, and accepted 17 (16% acceptance rate!) applications from 15 countries!. The list of accepted applications will be out soon.




Call for papers

We invite researchers to submit their recent work on Computer Vision applications, tasks and challenges inspired by and applied to developing regions, including: Paper presenters will be eligible for full (and partial) travel grants as well as oral presentations.
Researchers based in developing regions are strongly encouraged to submit.

Please note that, while there’s a wider range of work under the social good umbrella, we are particularly interested in works applied to and challenges coming from developing regions.

Submission Instructions: Papers should be submitted using the CVPR 2019 Latex/Word Templates and follow the CVPR formatting instructions. The review process is double-blind. The length must not exceed 6 pages (excluding references).

Submissions will be handled via CMT: https://cmt3.research.microsoft.com/CV4GC2019

The Camera Ready deadline was May 8th (midnight PST), 2019.




Call for Research Proposals

In partnership with the CV4GC workshop, Facebook AI is calling for research proposals that extend computer vision technologies to achieve global development priorities, especially those captured in the UN Sustainable Development Goals. Find more information about the call and apply through the Facebook AI website.

The deadline to apply has now passed.




Call for Computer Vision Ambassadors

Taking a global perspective on computer vision, also means widening participation. We will bring people who come to CVPR for the first time, which can be overwhelming (you probably still remember your first CVPR). We want these people to make the most of it, be prepared to showcase their material etc. We are therefore accepting volunteers from the CV community, that would act as "vision ambassadors" for each newcomer that needs help.

If accepted, you will be the mentor of a CVPR newcomer, and you are expected to connect with them before the conference, answer questions via email or video conferencing, assist them with their presentation and poster printing in needed. During the conference, you will help the newcomer orient, socialize and connect to the right people. We require that you have attended at least 2 top-tier Computer Vision conferences (CVPR/ICCV/ECCV) during the last couple of years to qualify.

Important Dates




Workshop Organizers

Lead Organizers
  • Laura Sevilla-Lara (University of Edinburgh & Facebook AI). Primary contact, email: [laura.sevilla.lara][at][gmail.com]
  • Yannis Kalantidis (Facebook AI). Primary contact, email: [yannisk][at][fb.com]
Challenges and Travel Lead
  • Drew Westbury (Facebook)
International Development Lead
  • Anna Lerner (Facebook)
Program Lead
  • Maria De-Arteaga (Carnegie Mellon University)
Computer Vision Ambassador program lead
  • Kris Sankaran (MILA)
Diversity Lead
  • Timnit Gebru (Google AI)
Sponsorship Leads
  • Julia Rhodes Davis (Partnership on AI)
  • Peter Eckersley (Partnership on AI)
  • Peter Eckersley (Partnership on AI)
Technical Program Chairs
  • Ernest Mwebaze (Google AI Ghana)
  • Mourad Gridach (Ibn Zohr University, Agadir, Morocco)
  • Stefano Ermon (Stanford University)
  • Lorenzo Torresani (Facebook AI)
Advisory Committee
  • Mutembesa Daniel (Kampala University, Uganda)
  • Rebekkah Hogan (Facebook Academic Relations)
  • John Quinn (Makerere University & Google AI Ghana)
  • Padmanabhan Anandan (Wadhwani AI, India)
  • Amir Zamir (UC Berkeley & Stanford University)
  • Larry Zitnick (University of Washington & Facebook Research)
  • Jitendra Malik (UC Berkeley & Facebook Research)
Computer Vision Ambassadors
We want to thank our amazing ambassadors for their help and contributions:
  • Adín Ramírez Rivera
  • Alejandro Galindo
  • Amanda Duarte
  • Anurag Arnab
  • Burak Uzkent
  • Cecilia Zhang
  • Ferda Ofli
  • Giorgos Tolias
  • Javier Romero
  • Jigar Doshi
  • Michela Paganini
  • Negar Rostamzadeh
  • Pulkit Agrawal
  • Qian Yu
  • Simone Schaub
  • Xavier Giro-i-Nieto
  • Zhongqi Miao

Sponsors

Platinum




Gold


Image Credit: NASA/GSFC/Jeff Schmaltz/MODIS Land Rapid Response Team