K-means Clustering Based Feature Consistency Alignment for Label-free Model Evaluation

TL;DR

This paper introduces a novel label-free model evaluation method using K-means clustering to align feature distributions and predict model performance without ground truth labels, demonstrating significant improvements in accuracy and robustness.

Contribution

We propose K-means Clustering Based Feature Consistency Alignment (KCFCA), a new approach for label-free model evaluation that handles distribution shifts and enhances performance prediction.

Findings

Secured 2nd place on DataCV Challenge leaderboard.

Achieved 36% improvement over baseline RMSE.

Produced more robust and optimal single model performance.

Abstract

The label-free model evaluation aims to predict the model performance on various test sets without relying on ground truths. The main challenge of this task is the absence of labels in the test data, unlike in classical supervised model evaluation. This paper presents our solutions for the 1st DataCV Challenge of the Visual Dataset Understanding workshop at CVPR 2023. Firstly, we propose a novel method called K-means Clustering Based Feature Consistency Alignment (KCFCA), which is tailored to handle the distribution shifts of various datasets. KCFCA utilizes the K-means algorithm to cluster labeled training sets and unlabeled test sets, and then aligns the cluster centers with feature consistency. Secondly, we develop a dynamic regression model to capture the relationship between the shifts in distribution and model accuracy. Thirdly, we design an algorithm to discover the outlier model factors, eliminate the outlier models, and combine the strengths of multiple autoeval models. On the DataCV Challenge leaderboard, our approach secured 2nd place with an RMSE of 6.8526. Our method significantly improved over the best baseline method by 36\% (6.8526 vs. 10.7378). Furthermore, our method achieves a relatively more robust and optimal single model performance on the validation dataset.