33rd International Conference on Computer-Aided Verification
Registration is now open. Register for CAV’21 here. Early registration is possible until July 9 2021, with regular registration being possible up to the conference. The registration fees for students are really low ($10-$30) and we offer NSF-funded fellowships for US students (application deadline is July 2). We hope this will encourage many students to join us this year at CAV. We are looking forward to seeing you at CAV 2021!
Participating in CAV’21
Note: After the conference, the CAV’21 conference page (cav21.org) has been made public and is accessible here without any registration.
If you have registered for CAV’21, an account has been created for you at cav21.org. It uses the email address you registered with. Please activate this account so you have access to the conference page as well as to the cav21 slack workspace and our gather. Make sure to read the Getting Started paragraph on cav21.org to learn more and find the links to slack and gather.
On cav21.org, you will find the relevant information for all events at CAV’21, including the zoom links. If you have any questions, feel free to ask in the helpdesk channel on slack (or send an email to firstname.lastname@example.org if you have trouble joining us there). Note that the slack workspace and the support address are not maintained anymore after the conference.
Virtual Conference Format
Please be aware that the following information is tentative and might change again.
Dates and Times: CAV will take place online Sunday July 18th – Saturday July 24th, with the main conference on Tuesday-Friday. Every day will be 4h of activities, 8am-12pm PDT (which is 4-8pm UK time). We will be mirroring the schedule 11 hours later at 8am-12pm in GMT+6 (which is 7-11pm PDT and 3-7am UK time), to enable wider participation of colleagues in Asia/Australia time zones. Workshops might follow a different schedule, information can be found on their websites.
Communication channels: We will be using Slack for (asynchronous) communication and discussions on papers and Gather for interactions with our sponsors and social events – join us for the pub quiz!
Program: We will have a virtual conference website which will contain the schedule and all relevant links. You can find the (tentative) publicly accessible schedulehere, other parts of the page are only accessible to CAV attendees.
Presentations: Each paper will have a 5min talk and a 25min talk. The 5min talk will be streamed during the conference and the 25min talk will be available before and after the conference for people to watch. So make sure to check out the talks already during the week before CAV!
Watch Parties: Feel free to organize (safe) watch parties with your fellow researchers and friends and let us know on social media (#cav21).
CAV Paper Sessions
Every session will be held twice: Once during the first block (8am-12pm PDT, 4-8pm UK time, 9pm-1am GMT+6) and then again in the second block 11 hours later (7-11pm PDT, 3-7am UK time, 8am-12pm GMT+6). Note that CAV will have parallel sessions. The following schedule is tentative and might change until the conference.
We give one day and time per session. For the first block, this is the time in PDT on that day. For the second block, this is the time in GMT+6 on the following day. As an example, a session given with 8-9 on July 20 will happen once at 8-9am PDT (which is 4-5pm UK time and 9-10pm GMT+6) and will be repeated at 8-9am on July 21 GMT+6 (which is 7-8pm on July 20 PDT and 3-4am on July 21 UK time). David Purser created a dynamic version of the following schedule that displays the times in your selected time zone, you can find it here.
Day 1 – July 20
Session 1 (9:20-10:20) – AI
DNNV: A Framework for Deep Neural Network Verification
Robustness Verification of Quantum Classifiers
BDD4BNN: A BDD-based Quantitative Analysis Framework for Binarized Neural Networks
Automated Safety Verification of Programs Invoking Neural Networks
Scalable Polyhedral Verification of Recurrent Neural Networks
Verisig 2.0: Verification of Neural Network Controllers Using Taylor Model Preconditioning
Robustness Verification of Semantic Segmentation Neural Networks using Relaxed Reachability