2019年CV 领域依然硕果累累，诞生了多篇<b>优秀论文</b>。我们选择了其中<b>十篇</b>论文，以供大家参考、学习，了解该领域的最新趋势与前沿技术。<br>
这十篇论文涵盖了卷积网络的优化，计算机视觉中的无监督学习，图像生成和机器生成图像的评估，视觉语言导航，使用自然语言为两个图像标注变化等。<br>
1.EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks<b>EfficientNet：</b><b>卷积神经网络模型缩放的反思</b><br>
2.Learning the Depths of Moving People by Watching Frozen People<b>通过观看静止的人来学习移动的人的深度</b><br>
3.Reinforced Cross-Modal Matching and Self-Supervised Imitation Learning for Vision-Language Navigation<b>增强的跨模态匹配和自我监督的模仿学习，用于视觉语言导航</b><br>
4.A Theory of Fermat Paths for Non-Line-of-Sight Shape Reconstruction<b>非视线形状重构的费马路径理论</b><br>
5.Reasoning-RCNN: Unifying Adaptive Global Reasoning into Large-scale Object Detection<b>Reasoning-RCNN：</b><b>将自适应全局推理统一到大规模目标检测中</b><br>
6.Fixing the Train-Test Resolution Discrepancy<b>修复训练测试分辨率差异</b><br>
7.SinGAN: Learning a Generative Model from a Single Natural Image<b>SinGAN：</b><b>从单个自然图像中学习生成模型</b><br>
8.Local Aggregation for Unsupervised Learning of Visual Embeddings<b>视觉聚合的无监督学习的局部聚合</b><br>
9.Robust Change Captioning<b>强大的更改字幕</b><br>
10.HYPE: A Benchmark for Human eYe Perceptual Evaluation of Generative Models<b>HYPE：</b><b>人类对生成模型的 eYe 感知评估的基准</b>