Proper Issue Labeling
Proper issue labeling, crucial for efficient data processing and model training, focuses on accurately assigning labels to data points across diverse applications, from natural language processing to image recognition and autonomous driving. Current research emphasizes improving label accuracy through refined model architectures (e.g., transformers, convolutional neural networks), advanced algorithms (e.g., contrastive learning, reinforcement learning), and techniques to mitigate issues like data imbalance and non-IID data distributions. These advancements are significant because accurate labeling directly impacts model performance, interpretability, and the reliability of downstream applications, ultimately driving progress in various fields of artificial intelligence.
Papers
TDD-Bench Verified: Can LLMs Generate Tests for Issues Before They Get Resolved?
Toufique Ahmed, Martin Hirzel, Rangeet Pan, Avraham Shinnar, Saurabh Sinha
Who Walks With You Matters: Perceiving Social Interactions with Groups for Pedestrian Trajectory Prediction
Ziqian Zou, Conghao Wong, Beihao Xia, Qinmu Peng, Xinge You