Abstract
Deep learning-based appearance gaze estimation methods are gaining popularity due to their high accuracy and fewer constraints from the environment. However, existing high-precision models often rely on deeper networks, leading to problems such as large parameters, long training time, and slow convergence. In terms of this issue, this paper proposes a novel lightweight gaze estimation model FGI-Net(Fusion Global Information). The model fuses global information into the CNN, effectively compensating for the need of multi-layer convolution and pooling to indirectly capture global information, while reducing the complexity of the model, improving the model accuracy and convergence speed. To validate the performance of the model, a large number of experiments are conducted, comparing accuracy with existing classical models and lightweight models, comparing convergence speed with models of different architectures, and conducting ablation experiments. Experimental results show that compared with GazeCaps, the latest gaze estimation model, FGI-Net achieves a smaller angle error with 87.1% and 79.1% reduction in parameters and FLOPs, respectively (MPIIFaceGaze is 3.74°, EyeDiap is 5.15°, Gaze360 is 10.50° and RT-Gene is 6.02°). Moreover, compared with different architectural models such as CNN and Transformer, FGI-Net is able to quickly converge to a higher accuracy range with fewer iterations of training, when achieving optimal accuracy on the Gaze360 and EyeDiap datasets, the FGI-Net model has 25% and 37.5% fewer iterations of training compared to GazeTR, respectively.
Abstract (translated)
基于深度学习的外观注视估计方法因其高精度和环境约束较少而日益流行。然而，现有高精度模型往往依赖更深的网络，导致参数量大、训练时间长及收敛慢等问题。针对此问题，本文提出一种新型轻量级注视估计模型FGI-Net（融合全局信息）。该模型将全局信息融入CNN中，有效弥补了通过多层卷积与池化间接捕获全局信息的需求，同时降低了模型复杂度，提升了模型精度与收敛速度。为验证模型性能，本文进行了大量实验：与现有经典模型及轻量级模型对比精度，与不同架构模型对比收敛速度，并开展消融实验。实验结果表明，相较于最新注视估计模型GazeCaps，FGI-Net在参数量与计算量上分别减少87.1%与79.1%，同时实现更小的角度误差（MPIIFaceGaze数据集为3.74°，EyeDiap为5.15°，Gaze360为10.50°，RT-Gene为6.02°）。此外，在与CNN、Transformer等不同架构模型的对比中，FGI-Net能以更少的训练迭代次数快速收敛至更高精度区间——在Gaze360与EyeDiap数据集达到最优精度时，其训练迭代次数较GazeTR分别减少25%与37.5%。
URL
https://arxiv.org/abs/2411.18064
