Mixup provides interpolated training samples and allows the model to obt...
Driven by improved architectures and better representation learning
fram...
The tracking-by-detection paradigm today has become the dominant method ...
Scaling object taxonomies is one of the important steps toward a robust
...
Test-time adaptation (TTA) has attracted significant attention due to it...
Universal Domain Adaptation aims to transfer the knowledge between the
d...
Recently, memory-based approaches show promising results on semi-supervi...
Unsupervised domain adaptation (UDA) for semantic segmentation has been
...
When the trained physician interprets medical images, they understand th...
Unsupervised Domain Adaptation (UDA) for semantic segmentation has been
...
Unsupervised Domain Adaptation for semantic segmentation has gained imme...
Temporal correspondence - linking pixels or objects across frames - is a...
Recently, deep self-training approaches emerged as a powerful solution t...
Pursuing a more coherent scene understanding towards real-time vision
ap...
Panoptic segmentation has become a new standard of visual recognition ta...
Visual storytelling is a task of creating a short story based on photo
s...
In this paper, we investigate the problem of unpaired video-to-video
tra...
We present a simple yet effective prediction module for a one-stage dete...
We propose a novel feed-forward network for video inpainting. We use a s...
Blind video decaptioning is a problem of automatically removing text ove...
Video inpainting aims to fill spatio-temporal holes with plausible conte...
In this paper, we address the problem of unsupervised video summarizatio...
Objects and their relationships are critical contents for image
understa...
We propose Convolutional Block Attention Module (CBAM), a simple yet
eff...
Recent advances in deep neural networks have been developed via architec...
One-stage object detectors such as SSD or YOLO already have shown promis...