Semi-supervised Multimodal Hashing
This addresses the challenge of reducing manual labeling effort for multimodal retrieval in applications like social media, though it is incremental as it builds on existing hashing techniques.
The paper tackles the problem of retrieving nearest neighbors across correlated multimodal data, such as image-text pairs, by proposing a semi-supervised multimodal hashing method that uses fuzzy logic to estimate labels for unlabeled data, achieving performance comparable to supervised methods with 90% labels when using only 50% labels.
Retrieving nearest neighbors across correlated data in multiple modalities, such as image-text pairs on Facebook and video-tag pairs on YouTube, has become a challenging task due to the huge amount of data. Multimodal hashing methods that embed data into binary codes can boost the retrieving speed and reduce storage requirement. As unsupervised multimodal hashing methods are usually inferior to supervised ones, while the supervised ones requires too much manually labeled data, the proposed method in this paper utilizes a part of labels to design a semi-supervised multimodal hashing method. It first computes the transformation matrices for data matrices and label matrix. Then, with these transformation matrices, fuzzy logic is introduced to estimate a label matrix for unlabeled data. Finally, it uses the estimated label matrix to learn hashing functions for data in each modality to generate a unified binary code matrix. Experiments show that the proposed semi-supervised method with 50% labels can get a medium performance among the compared supervised ones and achieve an approximate performance to the best supervised method with 90% labels. With only 10% labels, the proposed method can still compete with the worst compared supervised one.