We propose an efficient approach to train large diffusion models with ma...
Diffusion models have emerged as a key pillar of foundation models in vi...
Cascaded computation, whereby predictions are recurrently refined over
s...
Diffusion models have recently emerged as a powerful framework for gener...
We propose Image-to-Image Schrödinger Bridge (I^2SB), a new class of
con...
Diffusion models have found widespread adoption in various areas. Howeve...
Large-scale diffusion-based generative models have led to breakthroughs ...
While modern machine learning models rely on increasingly large training...
Denoising diffusion models (DDMs) have shown promising results in 3D poi...
Denoising diffusion models (DDMs) have emerged as a powerful class of
ge...
Molecular complexes formed by proteins and small-molecule ligands are
ub...
Adversarial purification refers to a class of defense methods that remov...
A wide variety of deep generative models has been developed in the past
...
We introduce AdaViT, a method that adaptively adjusts the inference cost...
Score-based generative models (SGMs) have demonstrated remarkable synthe...
Although machine learning models trained on massive data have led to
bre...
Controllable generation is one of the key requirements for successful
ad...
Score-based generative models (SGMs) have recently demonstrated impressi...
Detecting out-of-distribution (OOD) samples plays a key role in open-wor...
Training deep neural networks requires gradient estimation from data bat...
Variational autoencoders (VAEs) are one of the powerful likelihood-based...
Energy-based models (EBMs) have recently been successful in representing...
Normalizing flows, autoregressive models, variational autoencoders (VAEs...
Phrase grounding, the problem of associating image regions to caption wo...
How far apart are two neural networks? This is a foundational question i...
Neural architecture search (NAS) aims to discover network architectures ...
Domain shift is unavoidable in real-world applications of object detecti...
The representation of the posterior is a critical aspect of effective
va...
Building a large image dataset with high-quality object masks for semant...
In many applications we seek to maximize an expectation with respect to ...
Boltzmann machines are powerful distributions that have been shown to be...
Training of discrete latent variable models remains challenging because
...
Collecting large training datasets, annotated with high-quality labels, ...
In this paper we present an approach for classifying the activity perfor...
Rich semantic relations are important in a variety of visual recognition...
We present a novel approach for discovering human interactions in videos...
Many visual recognition problems can be approached by counting instances...