Visual Data Diagnosis and Debiasing with Concept Graphs

Authors: Rwiddhi Chakraborty, Yinong O Wang, Jialu Gao, Runkai Zheng, Cheng Zhang, Fernando D De la Torre

NeurIPS 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experiments show that data augmentation based on a balanced concept distribution augmented by CONBIAS improves generalization performance across multiple datasets compared to state-of-the-art methods.
Researcher Affiliation Academia 1Carnegie Mellon University 2Ui T The Arctic University of Norway 3Texas A&M University
Pseudocode Yes Algorithm 1 Rebalance Sampling
Open Source Code Yes Code: https://github.com/rwchakra/conbias
Open Datasets Yes We use three datasets in our work: Waterbirds [45], Urban Cars [22], and COCO-GB [50], that are commonly used in the bias mitigation domain.
Dataset Splits Yes In Table 5 we present the train, validation, and test splits for our three datasets. Waterbirds 4795 1199 5794 Urban Cars 8000 1000 1000 COCO-GB 32582 1331 1000
Hardware Specification Yes We trained all models on a single NVIDIA RTX A4000
Software Dependencies No The paper mentions 'Py Torch [35]' but does not provide a specific version number for it or other software dependencies.
Experiment Setup Yes Following previous work, we use validation loss based checkpointing to choose the best model, the Adam optimizer with a learning rate of 10 3, a weight decay of 10 5, and a cosine learning schedule over 100 epochs.