To bridge this crucial gap, first, this paper presents a comprehensive benchmark suite to evaluate the out-of-distribution (OOD) performance of image compression methods.

In this work, we study the effect of occlusion on video action recognition.

In addition, we introduce a simple baseline, "RDumb", that periodically resets the

We refer to them astarget-causing confounders.

Models like LASSO and ridge regression are extensively used in practice due to their interpretability, ease of use, and strong theoretical guarantees.

Motivated bythesuccess ofdiffusion models inhigh-fidelity and photorealistic image generation, generative data augmentation is an emerging application of diffusion models.