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.

Gradual Fusion Transformer (GraFT) is proposed for multimodal Object Re-Identification (ReID), employing learnable fusion tokens to capture modality-specific and object-specific features, optimizing the ReID feature embedding space through a novel training paradigm combined with an augmented triplet loss, and demonstrating superior performance over established benchmarks while offering model size and performance balance with integrated neural network pruning.