Learning with Limited Labeled Data

Sample Augmentation / Synthesis

• Delta-encoder: an effective sample synthesis method for few-shot object recognition (NeurIPS 2018)

  • [code]

• LaSO: Label-Set Operations networks for multi-label few-shot learning (CVPR 2019)

  • [code]

• Jointly Optimize Data Augmentation and Network Training: Adversarial Data Augmentation in Human Pose Estimation (CVPR 2018)

  • [code]

Transfer Learning 

• Predict to Learn (arXiv 2019)

• SpotTune: Transfer Learning through Adaptive Fine-tuning (CVPR 2019)

  • [code]

• Co-regularized Alignment for Unsupervised Domain Adaptation (NeurIPS 2018)

  • [medium]

Few Shot Learning

• RepMet: Representative-based metric learning for classification and one-shot object detection (CVPR 2019)

  • [code]

• Baby steps towards few-shot learning with multiple semantics (arXiv 2019)

Back Home