GraFT: Gradual Fusion Transformer for Multimodal Re-Identification
This work addresses the problem of efficient multimodal representation learning for object re-identification, which is incremental as it builds on existing fusion methods.
The paper tackles the scalability limitations of multimodal object re-identification models by introducing the Gradual Fusion Transformer (GraFT), which uses learnable fusion tokens and a novel training paradigm to achieve state-of-the-art performance on established benchmarks.
Object Re-Identification (ReID) is pivotal in computer vision, witnessing an escalating demand for adept multimodal representation learning. Current models, although promising, reveal scalability limitations with increasing modalities as they rely heavily on late fusion, which postpones the integration of specific modality insights. Addressing this, we introduce the \textbf{Gradual Fusion Transformer (GraFT)} for multimodal ReID. At its core, GraFT employs learnable fusion tokens that guide self-attention across encoders, adeptly capturing both modality-specific and object-specific features. Further bolstering its efficacy, we introduce a novel training paradigm combined with an augmented triplet loss, optimizing the ReID feature embedding space. We demonstrate these enhancements through extensive ablation studies and show that GraFT consistently surpasses established multimodal ReID benchmarks. Additionally, aiming for deployment versatility, we've integrated neural network pruning into GraFT, offering a balance between model size and performance.