In this paper, we propose a sentiment-enriched lightweight network SeLiNet
and an end-to-end on-device pipeline for contextual emotion recognition in
images. SeLiNet model consists of body feature extractor, image aesthetics
feature extractor, and learning-based fusion network which jointly estimates
discrete emotion and human sentiments tasks. On the EMOTIC dataset, the
proposed approach achieves an Average Precision (AP) score of 27.17 in
comparison to the baseline AP score of 27.38 while reducing the model size by
>85%. In addition, we report an on-device AP score of 26.42 with reduction in
model size by >93% when compared to the baseline.

在本文中，我们提出了一个情感丰富的轻量级网络 SeLiNet 和一个端到端的设备上的上下文情感识别流程，SeLiNet 模型由主体特征提取器、图像美学特征提取器和学习融合网络组成，共同估计离散情绪和人类情感任务。在 EMOTIC 数据集上，所提出的方法在与基准 AP 得分 27.38 相比下取得了 27.17 的平均准确率（AP）得分，同时模型大小缩小了 85% 以上。此外，与基准相比，我们在设备上报告了 26.42 的 AP 得分，模型大小缩小了 93% 以上。