Fusion and Orthogonal Projection for Improved Face-Voice Association
This work addresses a specific challenge in multimodal learning for computer vision, offering an incremental improvement in efficiency and effectiveness for face-voice association tasks.
The paper tackles the problem of learning associations between faces and voices by proposing a fusion and orthogonal projection (FOP) mechanism to create a discriminative joint embedding space, achieving favorable performance against state-of-the-art methods on the VoxCeleb dataset for cross-modal verification and matching tasks.
We study the problem of learning association between face and voice, which is gaining interest in the computer vision community lately. Prior works adopt pairwise or triplet loss formulations to learn an embedding space amenable for associated matching and verification tasks. Albeit showing some progress, such loss formulations are, however, restrictive due to dependency on distance-dependent margin parameter, poor run-time training complexity, and reliance on carefully crafted negative mining procedures. In this work, we hypothesize that enriched feature representation coupled with an effective yet efficient supervision is necessary in realizing a discriminative joint embedding space for improved face-voice association. To this end, we propose a light-weight, plug-and-play mechanism that exploits the complementary cues in both modalities to form enriched fused embeddings and clusters them based on their identity labels via orthogonality constraints. We coin our proposed mechanism as fusion and orthogonal projection (FOP) and instantiate in a two-stream pipeline. The overall resulting framework is evaluated on a large-scale VoxCeleb dataset with a multitude of tasks, including cross-modal verification and matching. Results show that our method performs favourably against the current state-of-the-art methods and our proposed supervision formulation is more effective and efficient than the ones employed by the contemporary methods.