arXiv · cs.AI · Abiram Panchalingam, Indu P. Bodala, Stuart E. Middleton · 10 小时前模型评分32
DiTGaze：几何条件扩散增强实现少样本视线估计
DiTGaze: Geometry-Conditioned Diffusion Augmentation for Few-Shot Gaze Estimation
导读
DiTGaze 用几何条件潜在扩散 Transformer（DiT，Diffusion Transformer）改造双流 3D Gaussian Splatting 的卷积渲染阶段，在 ETH-XGaze 上把生成器级视线误差较 GazeGaussian 降低 4.1%。
分类置信度
中 · 待人工复核
- 只有单一媒体来源，需要人工复核。
来源级别：媒体与个人
事件类型：研究发布
证据摘录
- “We present DiTGaze, a calibration-efficient augmentation system that follows a two-stream 3D Gaussian Splatting backbone and reformulates its convolutional rendering stage as a geometry-conditioned latent Diffusion Transformer (DiT).”
- “Accepted at WACV 2027.”
- “Updated and extended version of the original 2025 preprint, including revised methodology and expanded few-shot gaze-estimation experiments”
来源：arXiv · cs.AI · arxiv.org