解讀計算機視覺的2020 ：自監督學習、數據跨域分析成研究熱點，疫情相關應用落地加快

{"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“解讀 2020”年終技術盤點系列文章之一。"}]}]},{"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，嚴重擾亂了原有的生活秩序，給各個行業帶來嚴重的打擊。在這一年裏，人工智能相關技術和產業也迎來了全新的問題和挑戰。"}]},{"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":"筆者將從計算機視覺在學術界的發展和工業界的應用兩個角度出發，對這一年來湧現出的新技術和新方法進行回顧和總結，並展望明年的發展趨勢。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":2},"content":[{"type":"text","text":"計算機視覺技術在學術界的發展"}]},{"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":"過去的幾年裏，科研人員主要着眼於計算機視覺技術在圖像分類、目標檢測以及語意分割等任務中的研究和應用。"}]},{"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":"隨着算法精度的不斷提升，自從今年開始，科研人員逐漸將注意力轉向自監督學習、數據跨域分析以及輕量化模型設計等更加基礎的研究工作中。"}]},{"type":"paragraph","attrs":{"indent":0,"number":0,"align":null,"origin":null}},{"type":"heading","attrs":{"align":null,"level":3},"content":[{"type":"text","text":"1.1 基於Transformer的目標檢測"}]},{"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）是今年備受關注的一項代表性工作。該工作摒棄了現有的經典目標檢測算法如SSD、YOLO、Faster-RCNN中廣泛使用的Anchor、NMS等處理過程，借鑑自然語言處理領域廣泛使用的Transformer結構來進行目標檢測任務，實現了真正意義上的“端到端目標檢測”，也該領域提供了一種全新的解決方案。"}]}]}