Abstract:
Multi-modal hashing, a promising method for scaling cross-view retrieval, is gaining attention as multimedia data grows exponentially. Most multi-modal hashing methods either unnaturally split the learning process into two stages or treat the discrete optimization problem as a continuous one, resulting in suboptimal results. Recently, a few discrete multi-modal hashing methods have tried to address such issues, but they still ignore important discrete constraints like hash bit balance and decorrelation. We propose “Enhanced Discrete Multi-modal Hashing (EDMH)” to learn binary codes and hashing functions simultaneously from the pairwise similarity matrix of data under discrete constraints. Since its optimization subproblems have closed-form solutions after introducing a couple of auxiliary variables, we can develop a fast iterative learning algorithm for EDMH, even though it looks more complex than other multi-modal hashing models. Our experimental results on three real-world datasets showed the usefulness of those previously ignored discrete constraints and showed that EDMH outperforms state-of-the-art competitors in several retrieval metrics and runs faster than most of them.
Note: Please discuss with our team before submitting this abstract to the college. This Abstract or Synopsis varies based on student project requirements.
Did you like this final year project?
To download this project Code with thesis report and project training... Click Here
