Timezone: »

Distribution Aligning Refinery of Pseudo-label for Imbalanced Semi-supervised Learning
Jaehyung Kim · Youngbum Hur · Sejun Park · Eunho Yang · Sung Ju Hwang · Jinwoo Shin

Thu Dec 10 09:00 AM -- 11:00 AM (PST) @ Poster Session 5 #1331

While semi-supervised learning (SSL) has proven to be a promising way for leveraging unlabeled data when labeled data is scarce, the existing SSL algorithms typically assume that training class distributions are balanced. However, these SSL algorithms trained under imbalanced class distributions can severely suffer when generalizing to a balanced testing criterion, since they utilize biased pseudo-labels of unlabeled data toward majority classes. To alleviate this issue, we formulate a convex optimization problem to softly refine the pseudo-labels generated from the biased model, and develop a simple algorithm, named Distribution Aligning Refinery of Pseudo-label (DARP) that solves it provably and efficiently. Under various class imbalanced semi-supervised scenarios, we demonstrate the effectiveness of DARP and its compatibility with state-of-the-art SSL schemes.

Author Information

Jaehyung Kim (KAIST)
Youngbum Hur (Samsung Advanced Institute of Technology)
Sejun Park (KAIST)
Eunho Yang (Korea Advanced Institute of Science and Technology; AItrics)
Sung Ju Hwang (KAIST, AITRICS)
Jinwoo Shin (KAIST)

More from the Same Authors