Cluster-Aware Prompt Ensemble Learning for Few-Shot Vision-Language Model Adaptation
This work addresses a specific bottleneck in few-shot adaptation of vision-language models for researchers and practitioners, offering an incremental improvement over existing prompt ensembling methods.
The paper tackles the problem of suboptimal performance in vision-language models due to feature averaging in prompt ensembling, proposing a cluster-aware framework that improves classification by aligning with visual feature distributions and achieves state-of-the-art results, e.g., boosting accuracy by up to 3.2% on few-shot benchmarks.
Vision-language models (VLMs) such as CLIP achieve zero-shot transfer across various tasks by pre-training on numerous image-text pairs. These models often benefit from using an ensemble of context prompts to represent a class. Despite being effective, conventional prompt ensembling that averages textual features of context prompts often yields suboptimal results. This is because feature averaging shifts the class centroids away from the true class distribution. To address this issue, we propose the Cluster-Aware Prompt Ensemble Learning (CAPEL) framework, which preserves the cluster nature of context prompts. CAPEL classifies images into one of several class clusters, each represented by a distinct prompt. Instead of ensembling prompts in the feature space, we perform ensembling in the classification logits space, aligning better with the visual feature distribution. To further optimize prompt fine-tuning while maintaining cluster-specific discriminative power, we introduce a cluster-preserving regularization term. This ensures that prompts remain distinct and specialized for different clusters, preventing collapse into a uniform direction. Additionally, we integrate an adaptive prompt weighting technique to dynamically adjust the attention weights for flawed or ambiguous prompts, ensuring robust performance across diverse datasets and tasks.