Long-tailed distributions frequently emerge in real-world data, where a large number of
minority categories contain a limited number of samples. Such imbalance issue …

Deep models trained on long-tailed datasets exhibit unsatisfactory performance on tail
classes. Existing methods usually modify the classification loss to increase the learning …

Distant supervision tackles the data bottleneck in NER by automatically generating training
instances via dictionary matching. Unfortunately, the learning of DS-NER is severely …

We focus on the confounding bias between language and location in the visual grounding
pipeline, where we find that the bias is the major visual reasoning bottleneck. For example …

Label distributions in real-world are oftentimes long-tailed and imbalanced, resulting in
biased models towards dominant labels. While long-tailed recognition has been extensively …

Unsupervised semantic segmentation aims to achieve high-quality semantic grouping
without human-labeled annotations. With the advent of self-supervised pre-training, various …

As a common problem in the visual world, contextual bias means the recognition may
depend on the co-occurrence context rather than the objects themselves, which is even …

Existing methods for video question answering (VideoQA) often suffer from spurious
correlations between different modalities, leading to a failure in identifying the dominant …

The visual world naturally exhibits a long-tailed distribution of open classes, which poses
great challenges to modern visual systems. Existing approaches either perform class re …

Many objects do not appear frequently enough in complex scenes (eg, certain handbags in
living rooms) for training an accurate object detector, but are often found frequently by …