Machine learning research with positive real-world impact.
Ambitious applied research, positive outcomes
Recent highlights
Our research is supported by access to massive datasets, close collaboration with world renowned academic faculty, and a uniquely scalable machine learning platform.
CVPR 2024 | Data-Efficient Multimodal Fusion on a Single GPU
CVPR 2024 | Data-Efficient Multimodal Fusion on a Single GPU
Abstract
The goal of multimodal alignment is to learn a single latent space that is shared between multimodal inputs. The most powerful models in this space have been trained using massive datasets of paired inputs and large-scale computational resources, making them prohibitively expensive to train in many practical scenarios. We surmise that existing unimodal encoders pre-trained on large amounts of unimodal data should provide an effective bootstrap to create multimodal models from unimodal ones at much lower costs. We therefore propose FuseMix, a multimodal augmentation scheme that operates on the latent spaces of arbitrary pre-trained unimodal encoders. Using FuseMix for multimodal alignment, we achieve competitive performance — and in certain cases outperform state-of-the art methods — in both image-text and audio-text retrieval, with orders of magnitude less compute and data: for example, we outperform CLIP on the Flickr30K text-to-image retrieval task with ∼600× fewer GPU days and ∼80× fewer image-text pairs. Additionally, we show how our method can be applied to convert pre-trained text-to-image generative models into audio-to-image ones.
ICLR 2024 | Self-supervised Representation Learning from Random Data Projectors
ICLR 2024 | Self-supervised Representation Learning from Random Data Projectors
Abstract
Self-supervised representation learning (SSRL) has advanced considerably by exploiting the transformation invariance assumption under artificially designed data augmentations. While augmentation-based SSRL algorithms push the boundaries of performance in computer vision and natural language processing, they are often not directly applicable to other data modalities, and can conflict with application-specific data augmentation constraints. This paper presents an SSRL approach that can be applied to any data modality and network architecture because it does not rely on augmentations or masking. Specifically, we show that high-quality data representations can be learned by reconstructing random data projections. We evaluate the proposed approach on a wide range of representation learning tasks that span diverse modalities and real-world applications. We show that it outperforms multiple state-of-the-art SSRL baselines. Due to its wide applicability and strong empirical results, we argue that learning from randomness is a fruitful research direction worthy of attention and further study.
TMLR 2024 | Neural Implicit Manifold Learning for Topology-Aware Density Estimation
TMLR 2024 | Neural Implicit Manifold Learning for Topology-Aware Density Estimation
Abstract
Natural data is often constrained to an low-dimensional manifold. This work focuses on the task of building theoretically principled generative models for such data. Current generative models learn the manifold by mapping an low-dimensional latent variable through a neural network. These procedures, which we call pushforward models, incur a straightforward limitation: manifolds cannot in general be represented with a single parameterization, meaning that attempts to do so will incur either computational instability or the inability to learn probability densities within the manifold. To remedy this problem, we propose to model the manifold as a neural implicit manifold: the set of zeros of a neural network. We then learn the probability density within the manifold with a constrained energy-based model, which employs a constrained variant of Langevin dynamics to train and sample from the learned manifold. In experiments on synthetic and natural data, we show that our model can learn manifold-supported distributions with complex topologies more accurately than pushforward models.
NeurIPS 2023 | Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
NeurIPS 2023 | Exposing flaws of generative model evaluation metrics and their unfair treatment of diffusion models
Abstract
We systematically study a wide variety of image-based generative models spanning semantically-diverse datasets to understand and improve the feature extractors and metrics used to evaluate them. Using best practices in psychophysics, we measure human perception of image realism for generated samples by conducting the largest experiment evaluating generative models to date, and find that no existing metric strongly correlates with human evaluations. Comparing to 16 modern metrics for evaluating the overall performance, fidelity, diversity, and memorization of generative models, we find that the state-of-the-art perceptual realism of diffusion models as judged by humans is not reflected in commonly reported metrics such as FID. This discrepancy is not explained by diversity in generated samples, though one cause is over-reliance on Inception-V3. We address these flaws through a study of alternative self-supervised feature extractors, find that the semantic information encoded by individual networks strongly depends on their training procedure, and show that DINOv2-ViT-L/14 allows for much richer evaluation of generative models. Next, we investigate data memorization, and find that generative models do memorize training examples on simple, smaller datasets like CIFAR10, but not necessarily on more complex datasets like ImageNet. However, our experiments show that current metrics do not properly detect memorization; none in the literature is able to separate memorization from other phenomena such as underfitting or mode shrinkage. To facilitate further development of generative models and their evaluation we release all generated image datasets, human evaluation data, and a modular library to compute 16 common metrics for 8 different encoders.
NeurIPS 2023 | Adversarially robust learning with uncertain perturbation sets
NeurIPS 2023 | Adversarially robust learning with uncertain perturbation sets
Abstract
In many real-world settings exact perturbation sets to be used by an adversary are not plausibly available to a learner. While prior literature has studied both scenarios with completely known and completely unknown perturbation sets, we propose an in-between setting of learning with respect to a class of perturbation sets. We show that in this setting we can improve on previous results with completely unknown perturbation sets, while still addressing the concerns of not having perfect knowledge of these sets in real life. In particular, we give the first positive results for the learnability of infinite Littlestone classes when having access to a perfect-attack oracle. We also consider a setting of learning with abstention, where predictions are considered robustness violations, only when the wrong prediction is made within the perturbation set. We show there are classes for which perturbation-set unaware learning without query access is possible, but abstention is required.
Our research areas include recommendation systems, computer vision, time series forecasting, and natural language processing.
Big vision, deep roots
The co-founders of Layer 6, Jordan Jacobs and Tomi Poutanen, are also founders of the Vector Institute for Artificial Intelligence, and we maintain multiple research initiatives with Vector faculty. Current and former scientific advisors include professors Raquel Urtasun, Sanja Fidler, Rich Zemel, David Duvenaud, Laura Rosella and Scott Sanner.
Meaningful partnerships
Originally founded in 2011, Layer 6 now forms the AI research lab of TD Bank Group. Layer 6 impacts the lives of 25 million customers, helping more people achieve their financial goals. Partnerships with TD Securities provides Layer 6 with market data for training algo trading systems.
Layer 6 embraces opportunities to collaborate with Toronto’s world-leading medical research community, offering deep learning solutions to transform healthcare delivery and improve health outcomes. We are the first to deploy deep learning models on health data covering large population.
Passion to learn, driven to succeed
Our team represents 18 different countries of birth and we care deeply about fostering an inclusive culture where we learn from each other and win together. We are united by our passion for deep learning and a desire to apply our skills to have an outsized and positive impact on the future.
Meet some of our team
Develop your career at Layer 6
We’re growing our team exclusively with people driven to be at the top of the game in machine learning.
In the news
-
Inside TD’s AI play: How Layer 6’s technology hopes to improve old-fashioned banking advice
Globe And Mail
Read articleon Globe And Mail -
Tomi Poutanen: Geoffrey Hinton's Turing Award celebrates a life devoted to ground-breaking AI research
TD Newsroom
Read articleon TD Newsroom -
Tomi Poutanen: Chief Artificial Intelligence Officers Enter the C-Suite
Wall Street Journal
Read articleon Wall Street Journal -
TD Bank’s ‘Layer 6’ to bring machine learning personalization to diabetes care
IT Business
Read articleon IT Business -
TD Advances Innovation in Canadian Healthcare
TD Bank Group
Read articleon TD Bank Group
-
Inside TD’s AI play: How Layer 6’s technology hopes to improve old-fashioned banking advice
Globe And Mail
Read articleon Globe And Mail -
Tomi Poutanen: Geoffrey Hinton's Turing Award celebrates a life devoted to ground-breaking AI research
TD Newsroom
Read articleon TD Newsroom -
Tomi Poutanen: Chief Artificial Intelligence Officers Enter the C-Suite
Wall Street Journal
Read articleon Wall Street Journal -
TD Bank’s ‘Layer 6’ to bring machine learning personalization to diabetes care
IT Business
Read articleon IT Business -
TD Advances Innovation in Canadian Healthcare
TD Bank Group
Read articleon TD Bank Group -
Jordan Jacobs, co-founder of Vector Institute on Canada as a global AI leader
IT Business
Read articleon IT Business -
Layer 6’s Jordan Jacobs: Canada needs to promote itself as an AI leader
BetaKit
Read articleon BetaKit -
U of T alumni and graduate students part of Layer 6 AI's win in global competition
U of T News
Read articleon U of T News -
Tomi Poutanen and Michael Rhodes discuss the future of artificial intelligence with Amanda Lang
TD Bank Group
Read articleon TD Bank Group -
Get Smart – Artificial intelligence is transforming business and life
Ivey Business School
Read articleon Ivey Business School -
How TD Bank plans to use artificial intelligence
BNN Bloomberg
Read articleon BNN Bloomberg