Feature Alignment: Rethinking Efficient Active Learning via Proxy in the Context of Pre-trained Models
This work addresses the computational cost and performance trade-off in active learning for pre-trained models, offering an incremental improvement for researchers and practitioners in efficient machine learning.
The paper tackles the performance degradation in proxy-based active learning for fine-tuning pre-trained models by showing that not all sample selection differences cause degradation and that suitable training methods can mitigate it. They propose a method that updates pre-computed features and selects proper training, improving total cost while maintaining efficiency, as validated in experiments.
Fine-tuning the pre-trained model with active learning holds promise for reducing annotation costs. However, this combination introduces significant computational costs, particularly with the growing scale of pre-trained models. Recent research has proposed proxy-based active learning, which pre-computes features to reduce computational costs. Yet, this approach often incurs a significant loss in active learning performance, sometimes outweighing the computational cost savings. This paper demonstrates that not all sample selection differences result in performance degradation. Furthermore, we show that suitable training methods can mitigate the decline of active learning performance caused by certain selection discrepancies. Building upon detailed analysis, we propose a novel method, aligned selection via proxy, which improves proxy-based active learning performance by updating pre-computed features and selecting a proper training method. Extensive experiments validate that our method improves the total cost of efficient active learning while maintaining computational efficiency. The code is available at \url{https://github.com/ZiTingW/asvp}.