Adel Bibi

Senior Researcher

Biography

Adel Bibi is a senior researcher in machine learning and computer vision at the Department of Engineering Science of the University of Oxford, a Junior Research Fellow (JRF) at Kellogg College, and a member of the ELLIS Society. Prior to that, Bibi was a senior research associate and a postdoctoral researcher with Philip H.S. Torr since October 2020. He earned his MSc and PhD degrees from King Abdullah University of Science & Technology (KAUST) in 2016 and 2020, respectively, advised by Bernard Ghanem. Bibi was awarded an Amazon Research Award in 2022 in the Machine Learning Algorithms and Theory track. Bibi received four best paper awards: a NeurIPS23 workshop, an ICML23 workshop, a 2022 CVPR workshop, and one at Optimization and Big Data Conference in 2018. His contributions include over 30 papers published in top machine learning and computer vision conferences. He also received four outstanding reviewer awards (CVPR18, CVPR19, ICCV19, ICLR22) and a notable Area Chair Award in NeurIPS23.

Currently, Bibi is interested in large scale offline and online robust and private continual learning. Robustness, in both aspects empirical and provably certifiable, here refers to deep models under $\ell_p$ bounded additive and geometric attacks. Moreover, continual learning refers to the learning from a stream of data in stringent memory and computational settings.

Download my resume

[Note!] I am always looking for strong self-motivated PhD students. If you are interested in Trustworthy Foundation Models that Continually Learn, reach out!

[Consulting Expertise] I have consulted in the past for projects spanning core machine learning and data science, computer vision, certification and AI safety, optimization formulations for matching and resource allocation problems, among other problems.

Interests

Trustworthy AI and Safety
Robustness and Certification
Continual Learning
Optimization

Education

PhD in Electrical Engineering (4.0/4.0); Machine Learning and Optimization Track, 2020

King Abdullah University of Science and Technology (KAUST)
MSc in Electrical Engineering (4.0/4.0); Computer Vision Track, 2016

King Abdullah University of Science and Technology (KAUST)
BSc in Electrical Engineering (3.99/4.0), 2014

Kuwait University

Officially Co-supervised PhD Students

Francisco Eiras

PhD Student, University of Oxford

Csaba Botos

PhD Student, University of Oxford

Cornelius Emde

PhD Student, University of Oxford

Ameya Prabhu

PhD Student, University of Oxford

Aleksandar Petrov

PhD Student, University of Oxford

News

[February 15th, 2024]: I gave a talk to TAHAKOM on AI Safety Research.
[February 13th, 2024]: I gave a talk at the AI Safety meeting in the Said Busienss School in Oxford in our work on AI Safety.
[February 7th, 2024]: Our paper SynthCLIP: Are We Ready for a Fully Synthetic CLIP Training? on training CLIP with only synthetic data was advertised on TLDR News which has over 1M subscribers.
[January 21th, 2024]: Three papers accepted to ICLR24 (one as spotlight).

~~ End of 2023 ~~

[December 16th, 2023]: Our paper When Do Prompting and Prefix-Tuing Work? A Theory of Capabilities and imitations received the Entropic Paper Award at the NeurIPS 2023 Workshop: I Can’t Believe It’s Not Better (ICBINB): Failure Modes in the Age of Foundation Models.
[December 12th, 2023]: Received a Notable Area Chair Award in NeurIPS23 (awarded to 8.1% of 1223).
[December 9th, 2023]: One paper (SimCS: Simulation for Domain Incremental Online Continual Segmentation) accepted to AAAI24.
[November 2nd, 2023]: I gave two talks on trustwrothy AI to Kuwait University and Intematix, a startup based in Riyadh in Saudi Arabia.
[September 21st, 2023]: One paper on Language Model Tokenizers Introduce Unfairness Between Languages has been accepted to NeurIPS23.
[September 5th, 2023]: Invited by Dima Damen to give a talk at Bristol University.
[July 18th, 2023]: One robustness paper is accepted to TMLR.
[July 11th, 2023]: Our paper on Provably Correct Physics-Informed Neural Networks received the outstanding paper award in the Formal Verification Workshop in ICML23.
[July 14th, 2023]: One paper on online continual learning is accepted to ICCV 2023.
[July 11th, 2023]: Our paper Provably Correct Physics-Informed Neural Networks received an outstanding paper award at WFVML ICML23 workshop.
[June 23rd, 2023]: One paper accepted to DeployableGenerativeAI (ICML23 Workshop).
[June 20th, 2023]: Three papers accepted to AdvML-Frontiers (ICML23 Workshop).
[June 18th, 2023]: Invited to give a talk at the CLVision workshop in CVPR23. I was also part of the panel discussion.
[April 24th, 2023]: One paper accepted to ICML 2023.
[March 29th, 2023]: Four papers accepted to CLVision (CVPR23 Workshop).
[March 9th, 2023]: Media coverage by DOU, a Ukrainian development community focused on technologies, frameworks, and code, for my work with Taras Rumezhak.
[March 1st, 2023]: I am promoted to a Senior Researcher (G9) in Machine Learning and Computer Vision at Oxford.
[February 28th, 2023]: I will serve as an Area Chair for the NeurIPS 2023.
[February 27th, 2023]: Two papers accepted to CVPR 2023. One of the two papers is accepted as a highlight (2.5% of ~9K submissions).
[January 23th, 2023]: I will serve as an Area Chair for the ICLR 2023 Workshop on Trustworthy ML.

~~ End of 2022 ~~

[December 9th, 2022]: I will serve as a Senior Program Committee (Area Chair/Meta Reviewer) for IJCAI 2023.
[November 3rd, 2022]: Joined the ELLIS Society as a member.
[October 11th, 2022]: One paper accepted to WACV23.
[October 10th, 2022]: Recieved an Amazon Research Award.
[September 14th, 2022]: Our paper N-FGSM accepted to NeurIPS22.
[August 28th, 2022]: Our paper ANCER was accepted in Transactions on Machine Learning Research (TMLR).
[August 15th, 2022]: Our paper on Tropical Geometry was accepted to appear in the IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI).
[July 11th, 2022]: I will serve as a Senior Program Committee (Area Chair/Meta Reviewer) for AAAI23.
[May 26th, 2022]: Three papers accepted to the Adversarial Machine Learning Frontiers ICML2022 workshop! Papers will be coming on arXiv soon.
[May 16th, 2022]: Our paper titled Data Dependent Randomzied Smoothing is accepted to UAI22.
[April 22nd, 2022]: I was selected as the Highlighted Reviewer of ICLR 2022 and received a free conference registration.

~~ End of 2021 ~~

[Dec 1st, 2021]: I got promoted to a Senior Research Associate (G8) in machine learning of the Torr Vision Group (TVG) at the University of Oxford.
[Dec 1st, 2021]: Two papers, Combating Adversaries with anti-adversaries and DeformRS: Certifying Input Deformations with Randomized Smoothing, are accepted in AAAI22.
[Nov 18th, 2021]: We were awarded KAUST’s Competitive Research Grant with a total of > 1.05M$ (one million USD). This is a 3 years collaboration between KAUST and Oxford.
[Oct 15th, 2021]: Rethinking Clustering for Robustness is accepted in BMVC21.
[June 22nd, 2021]: Anti Adversary paper is accepted in Adversarial Machine Learning Workshop @ICML21.
[June 13th, 2021]: I have been elected as a Junior Research Fellow of Kellogg College, University of Oxford. Appointment starts in October 2021.
[March 28th, 2021]: ETB robustness paper is accepted in RobustML Workshop @ICLR21.

~~ End of 2020 ~~

[November 8th, 2020]: New paper on robustness is on arXiv.
[November 8th, 2020]: New paper on randomized smoothing is on arXiv.
[October 15th, 2020]: I joined the Torr Vision Group working with Philip Torr at the University of Oxford.
[July 2nd, 2020]: Gabor layers enhance robustness paper accepted to ECCV20 arXiv.
[June 30th, 2020]: One paper is out on new expressions for the output moments of ReLU based networks with various new appliactions arXiv.
[June 24th, 2020]: New paper with SOTA results, backed with theory, on training robust models through feature clustering arXiv.
[March 31st, 2020]: I have sucessfully defended my PhD thesis.

~~ End of 2019 ~~

[Dec 20th, 2019]: One paper accepted to ICLR20.
[Nov 11th, 2019]: One spotlight paper accepted to AAAI20.
[Sept 25th, 2019]: Recognized as outstanding reviewer for ICCV19. Link.
[August 5th, 2019]: I was invited to give a talk about the most recent research in computer vision and machine learning from the IVUL group at PRIS19, Dead Sea, Jordan. I also gave a 1 hour long workshop about deep learning and pytorch. Slides1/Slides2/Material.
[July 6th, 2019]: I was invited to give a talk at the Eastern European Conference on Computer Vision, Odessa, Ukraine. Slides.
[June 28th, 2019]: I gave a talk at the Biomedical Computer Vision Group directed by Prof Pablo Arbelaez, Bogota, Colombia. Slides.
[June 15th, 2019]: Attended CVPR19.
[June 9th, 2019]: Recognized as an outstanding reviewer for CVPR19. This is the second time in a row for CVPR. Check it out. :)
[May 26th, 2019]: A new paper is out on derivative free optimization with momentum with new rates and results on continuous controls tasks. arXiv.
[May 25th, 2019]: New paper! New provably tight interval bounds are derived for DNNs. This allows for very simple robust training of large DNNs. arXiv.
[May 11th, 2019]: How to train robust networks outperforming 2-21x fold data augmentation? New paper out on arXiv.
[May 6th, 2019]: Attended ICLR19 in New Orleans.
[Feb 4th, 2019]: New paper on derivative-free optimization with importance sampling is out! Paper is on arXiv.

~~ End of 2018 ~~

[Dec 22nd, 2018]: One paper accepted to ICLR19, Louisiana, USA.
[Nov 6th, 2018]: One paper accepted to WACV19, Hawaii, USA.
[July 3rd, 2018]: One paper accepted to ECCV18, Munich, Germany.
[June 19th, 2018]: Attended CVPR18 and gave an oral talk on our most recent work on analyzing piecewise linear deep networks using Gaussian network moments. Tensorflow, Pytorch and MATLAB codes are released.
[June 17th, 2018]: Received a fully funded scholarship to attend the AI-DLDA 18 summer school in Udine, Italy. Unfortunately, I won’t be able to attend for time constraints. Link
[June 15th, 2018]: New paper out! “Improving SAGA via a Probabilistic Interpolation with Gradient Descent”.
[April 30th, 2018]: I’m interning for 6 months at the Intel Labs in Munich this summer with Vladlen Koltun.
[April 22nd, 2018]: Recognized as an outstanding reviewer for CVPR18. I’m also on the list of emergency reviewers. Check it out. :)
[March 6th, 2018]: One paper accepted as [Oral] in CVPR 2018.
[Feb 5, 2018]: Awarded the best KAUST poster prize in the Optimization and Big Data Conference.

~~ End of 2017 ~~

[Decemmber 11, 2017]: TCSC code is on github.
[October 22, 2017]: Attened ICCV17, Venice, Italy.
[July 22, 2017]: Attened CVPR17 in Hawaii and gave an oral presentation on our work on solving the LASSO with FFTs, July 2017.
[July 16, 2017]: FFTLasso’s code is available online.
[July 9, 2017]: Attended the ICVSS17, Sicily, Italy.
[June 15, 2017]: Selected to attend the International Computer Vision Summer School (ICVSS17), Sicily, Italy.
[March 17, 2017]: 1 paper accepted to ICCV17.
[March 14, 2017]: Received my NanoDegree on Deep Learning from Udacity.
[March 3, 2017]: 1 oral paper accepted to CVPR17, Hawai, USA.

~~ End of 2016 ~~

[October 19, 2016]: ECCV16’s [code](https://github.com/adelbibi/ Target-Response-Adaptation-for-Correlation-Filter-Tracking) has been released on github.
[October 8, 2016]: Attended ECCV16, Amsterdam, Netherlands.
[July 11, 2016]: 1 spotlight paper accepted to ECCV16, Amsterdam, Netherlands.
[June 26, 2016]: Attended CVPR16, Las Vegas, USA. Two papers presented.
[May 13, 2016]: ICCVW15 code is now avaliable online.
[April 11, 2016]: Successfully defended my Master’s Thesis.
[March 2, 2016]: 2 papers (1 spotlight) accepted to CVPR16, Las Vegas, USA.

~~ End of 2015 ~~

[November 20, 2015]: 1 paper acceted to ICCVW15, Santiago, Chile.
[June 8, 2015]: Attended CVPR15, Boston, USA.

Awards

Outstanding Paper Award; Entropic Paper Award in NeurIPS23 workshop

The award is presented to our paper “Paper: When Do Prompting and Prefix-Tuning Work? A Theory of Capabilities and Limitations” in the NeurIPS23 Workshop I Cant Believe Its Not Better (ICBINB): Failure Modes in the Age of Foundation Models.

Adel Bibi

Last updated on Jan 21, 2024 1 min read

Outstanding Paper Award; Entropic Paper Award in NeurIPS23 workshop

Notable Area Chair Award in NeurIPS23

Unfortunately, I did not attend NeurIPS 23 this year as I was ill (luckily, my student Aleksandar Petrov was there to present our paper). However, to my surprise, he sent me this image where I was selected as a notable Area Chair.

Adel Bibi

Last updated on Jan 21, 2024 1 min read

Outstanding Paper Award in ICML23 Workshop

The award is presented to our paper “Provably Correct Physics-Informed Neural Networks” in the Workshop on Formal Verification of Machine Learning (WFVML 2023) in ICML23. Congrats to the my student Francisco Eiras.

Adel Bibi

Last updated on Jan 21, 2024 1 min read

Outstanding Paper Award in ICML23 Workshop

ELLIS Member

I was nomiated to join the ELLIS the (European Laboratory for Learning Intelligence Systems). This is a network of AI European excellence focused on research around AI, machine learning, computer vision, and robotics.

Adel Bibi

Last updated on Jan 27, 2024 1 min read

Amazon Research Award 2022

I was awarded the Amazon Research Award in 2022 in the Machine Learning Algorithms and Theory track. The project proposal is “Randomized Smoothing; Future Directions and Extensions”

Adel Bibi

Last updated on Jan 21, 2024 2 min read

See all posts

Featured Publications

Aleksandra Petrov, Philip H.S. Torr, Adel Bibi

January, 2024 International Conference on Learning Representations (ICLR24)

When Do Prompting and Prefix-Tuning Work? A Theory of Capabilities and Limitations"

Context-based fine-tuning methods like prompting, in-context learning, soft prompting (prompt tuning) and prefix-tuning have gained popularity as they often match the performance of full fine-tuning with a fraction of the parameters. Despite their empirical successes, there is little theoretical understanding of how these techniques influence the internal computation of the model and their expressiveness limitations. We show that despite the continuous embedding space being much more expressive than the discrete token space, soft-prompting and prefix-tuning are strictly less expressive than full fine-tuning. Concretely, context-based fine-tuning cannot change the relative attention pattern over the content and can only bias the outputs of an attention layer in a fixed direction. While this means that fine-tuning techniques such as prompting, in-context learning, soft prompting and prefix-tuning can successfully elicit or combine skills already present in the pretrained model, they cannot learn tasks requiring new attention patterns.

Tim Franzmeyer, Stephen Marcus McAleer, Joao F. Henriques, Jakob Nicolaus Foerster, Philip H.S. Torr, Adel Bibi, Christian Schroeder de Witt

January, 2024 International Conference on Learning Representations (ICLR24) [Spotlight]

Illusory Attacks: Detectability Matters in Adversarial Attacks on Sequential Decision-Makers

Autonomous agents deployed in the real world need to be robust against adversarial attacks on sensory inputs. Robustifying agent policies requires anticipating the strongest attacks possible. We demonstrate that existing observation-space attacks on reinforcement learning agents have a common weakness; while effective, their lack of information-theoretic detectability constraints makes them detectable using automated means or human inspection. Detectability is undesirable to adversaries as it may trigger security escalations. We introduce illusory attacks, a novel form of adversarial attack on sequential decision-makers that is both effective and of bounded statistical detectability. We propose a novel dual ascent algorithm to learn such attacks end-to-end. Compared to existing attacks, we empirically find illusory attacks to be significantly harder to detect with automated methods, and a small study with human subjects suggests they are similarly harder to detect for humans. Our findings suggest the need for better anomaly detectors, as well as effective hardware and system-level defenses.

Wenxuan Zhang, Youssef Mohamed, Bernard Ghanem, Philip H. S. Torr, Adel Bibi, Mohamed Elhoseiny

January, 2024 International Conference on Learning Representations (ICLR24)

Continual Learning on a Diet: Learning from Sparsely Labeled Streams Under Constrained Computation

We propose and study a realistic Continual Learning (CL) setting where learning algorithms are granted a restricted computational budget per time step while training. We apply this setting to large-scale semi-supervised Continual Learning scenarios with sparse label rate. Previous proficient CL methods perform very poorly in this challenging setting. Overfitting to the sparse labeled data and insufficient computational budget are the two main culprits for such a poor performance. Our new setting encourages learning methods to effectively and efficiently utilize the unlabeled data during training. To that end, we propose a simple but highly effective baseline, DietCL, which utilizes both unlabeled and labeled data jointly. DietCL meticulously allocates computational budget for both types of data. We validate our baseline, at scale, on several datasets, e.g., CLOC, ImageNet10K, and CGLM, under constraint budget setup. DietCL outperforms, by a large margin, all existing supervised CL algorithms as well as more recent continual semi-supervised methods. Our extensive analysis and ablations demonstrate that DietCL is stable under a full spectrum of label sparsity, computational budget and various other ablations.

See all publications