TextAug: Test time Text Augmentation for Multimodal Person Re-identification
This addresses the bottleneck of needing large multimodal training datasets for researchers in computer vision and re-identification, though it is incremental as it adapts existing image augmentation methods to text.
The study tackled the challenge of augmenting text data in multimodal person re-identification by proposing CutMixOut, a test-time augmentation method combining cutout and cutmix techniques, which improved performance on multiple benchmarks without prior training.
Multimodal Person Reidentification is gaining popularity in the research community due to its effectiveness compared to counter-part unimodal frameworks. However, the bottleneck for multimodal deep learning is the need for a large volume of multimodal training examples. Data augmentation techniques such as cropping, flipping, rotation, etc. are often employed in the image domain to improve the generalization of deep learning models. Augmenting in other modalities than images, such as text, is challenging and requires significant computational resources and external data sources. In this study, we investigate the effectiveness of two computer vision data augmentation techniques: cutout and cutmix, for text augmentation in multi-modal person re-identification. Our approach merges these two augmentation strategies into one strategy called CutMixOut which involves randomly removing words or sub-phrases from a sentence (Cutout) and blending parts of two or more sentences to create diverse examples (CutMix) with a certain probability assigned to each operation. This augmentation was implemented at inference time without any prior training. Our results demonstrate that the proposed technique is simple and effective in improving the performance on multiple multimodal person re-identification benchmarks.