解读计算机视觉的2020 ：自监督学习、数据跨域分析成研究热点，疫情相关应用落地加快

{"type":"doc","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"blockquote","content":[{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"italic"},{"type":"strong"}],"text":"本文是 InfoQ“解读 2020”年终技术盘点系列文章之一。"}]}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"2020年，新冠疫情肆虐全球，严重扰乱了原有的生活秩序，给各个行业带来严重的打击。在这一年里，人工智能相关技术和产业也迎来了全新的问题和挑战。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","marks":[{"type":"strong"}],"text":"笔者将从计算机视觉在学术界的发展和工业界的应用两个角度出发，对这一年来涌现出的新技术和新方法进行回顾和总结，并展望明年的发展趋势。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"计算机视觉技术在学术界的发展"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"过去的几年里，科研人员主要着眼于计算机视觉技术在图像分类、目标检测以及语意分割等任务中的研究和应用。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"随着算法精度的不断提升，自从今年开始，科研人员逐渐将注意力转向自监督学习、数据跨域分析以及轻量化模型设计等更加基础的研究工作中。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"1.1 基于Transformer的目标检测"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null},"content":[{"type":"text","text":"Facebook提出的DETR（DEtection TRansformer）是今年备受关注的一项代表性工作。该工作摒弃了现有的经典目标检测算法如SSD、YOLO、Faster-RCNN中广泛使用的Anchor、NMS等处理过程，借鉴自然语言处理领域广泛使用的Transformer结构来进行目标检测任务，实现了真正意义上的“端到端目标检测”，也该领域提供了一种全新的解决方案。"}]}]}