GRMar 11

Landmark Guided 4D Facial Expression Generation

arXiv:2603.10337v112.04 citationsh-index: 60
Predicted impact top 40% in GR · last 90 daysOriginality Incremental advance
AI Analysis

This work addresses identity robustness in facial expression generation for computer vision applications, representing an incremental improvement over existing methods.

The paper tackles the problem of generating 4D facial expressions that are robust to identity changes by proposing LM-4DGAN, which uses neutral landmarks as guidance and incorporates an identity discriminator and landmark autoencoder into a WGAN framework, along with a cross-attention mechanism for identity-specific displacement decoding.

In this paper, we proposed a generative model that learns to synthesize the 4D facial expression with the neutral landmark. Existing works mainly focus on the generation of sequences guided by expression labels, speech, etc, while they are not robust to the change of different identities. Our LM-4DGAN utilizes neutral landmarks to guide the facial expression generation while adding an identity discriminator and a landmark autoencoder to the basic WGAN for achieving better identity robustness. Furthermore, we add a cross-attention mechanism to the existing displacement decoder which is suitable for the given identity.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes