Abstract:
Deep learning and computer vision researchers study interpretable data representation. Representation disentanglement is effective for this task, but existing works cannot easily handle problems that require manipulating and recognizing data across multiple domains.
We present a unified network architecture of Multi-domain and Multi-modal Representation Disentangler (M2RD) to learn domain-invariant content representation with the observed domain-specific representation.
The proposed model can continuously manipulate images across data domains and modalities using adversarial learning and disentanglement techniques.
Unsupervised domain adaptation is possible with the domain-invariant feature representation. Finally, our quantitative and qualitative results show that the proposed model is more effective and robust than state-of-the-art methods on the above tasks.
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
