Recently, deep learning-based facial landmark detection has achieved significant improvement. However, the semantic ambiguity problem degrades detection performance. Specifically, the semantic ambiguity causes inconsistent annotation and negatively affects the model's convergence, leading to worse accuracy and instability prediction. To solve this problem, we propose a Self-adapTive Ambiguity Reduction (STAR) loss by exploiting the properties of se-mantic ambiguity. We find that semantic ambiguity results in the anisotropic predicted distribution, which inspires us to use predicted distribution to represent semantic ambiguity. Based on this, we design the STAR loss that measures the anisotropism of the predicted distribution. Compared with the standard regression loss, STAR loss is encouraged to be small when the predicted distribution is anisotropic and thus adaptively mitigates the impact of semantic ambiguity. Moreover, we propose two kinds of eigen-value restriction methods that could avoid both distribution's abnormal change and the model's premature convergence. Finally, the comprehensive experiments demonstrate that STAR loss outperforms the state-of-the-art methods on three benchmarks, i.e., COFW, 300W, and WFLW, with negligible computation overhead. Code is at https://github.com/ZhenglinZhou/STAR

We present an online Item Intelligent Publishing System used in large scale Consumer to Consumer (C2C) transaction platform, named as I2PS, it is designed for personal seller to publish their items in an automatic way with the uploaded images. Our I2PS is deployed in Xianyu mobile App, the largest second-hand item shopping platform in China. The proposed system not only can guide seller how to photograph the publishing items with more details based on category recognition module, but also can intelligently tell the seller exactly what this product is and which attributes it has based on various recognition methods. The seller does not need to input extra product information, so that the item's publish process in Xianyu can be performed without any difficulty. In this paper, we introduce several techniques we used to develop the I2PS for product understanding, including product's category recognition, Standard Product Unit (SPU) recognition, multi-label attribute recognition and their corresponding pre-processing technologies. Our system deployed in Xianyu can help tens of millions personal sellers to publish their items, and improves publishing success rate by more than 15% and reduces publishing duration by more than 20%. The demo video is available at https://youtu.be/3NRx2hECIHc.

In this paper, we propose an end-to-end deep learning architecture that generates 3D triangular meshes from single color images. Restricted by the nature of prevalent deep learning techniques, the majority of previous works represent 3D shapes in volumes or point clouds. However, it is non-trivial to convert these representations to compact and ready-to-use mesh models. Unlike the existing methods, our network represents 3D shapes in meshes, which are essentially graphs and well suited for graph-based convolutional neural networks. Leveraging perceptual features extracted from an input image, our network produces the correct geometry by progressively deforming an ellipsoid. To make the whole deformation procedure stable, we adopt a coarse-to-fine strategy, and define various mesh/surface related losses to capture properties of various aspects, which benefits producing the visually appealing and physically accurate 3D geometry. In addition, our model by nature can be adapted to objects in specific domains, e.g., human faces, and be easily extended to learn per-vertex properties, e.g., color. Extensive experiments show that our method not only qualitatively produces the mesh model with better details, but also achieves the higher 3D shape estimation accuracy compared against the state-of-the-arts.