Ashish Patel 🇮🇳’s Post

𝗗𝗮𝘆-𝟭𝟴𝟮 Computer Vision Learning 𝗛𝗢𝗧𝗥: End-to-End Human-Object Interaction Detection with Transformers by Korea University and 카카오브레인 - kakaobrain Follow me for a similar post:  🇮🇳 Ashish Patel Interesting Facts : 🔸 This is a paper in CVPR 2021 with over 13 citations. 🔸 It outperforms IPNet, UnionDet, etc. ------------------------------------------------------------------- 𝗔𝗺𝗮𝘇𝗶𝗻𝗴 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 : https://lnkd.in/eQ7ZXDW Code : https://lnkd.in/epVAvjG ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 🔸 Human-Object Interaction (HOI) detection is a task of identifying “a set of interactions” in an image, which involves the i) localization of the subject (\ie, humans) and target (\ie, objects) of interaction, and ii) the classification of the interaction labels.  🔸 Most existing methods have indirectly addressed this task by detecting human and object instances and individually inferring every pair of the detected instances. 🔸 In this paper, the authors present a novel framework, referred to by HOTR, which directly predicts a set of human, object, interaction triplets from an image based on a transformer encoder-decoder architecture. 🔸 Through the set prediction, our method effectively exploits the inherent semantic relationships in an image and does not require time-consuming post-processing which is the main bottleneck of existing methods. The proposed algorithm achieves the state-of-the-art performance in two HOI detection benchmarks with an inference time under 1 ms after object detection. #computervision #artificialintelligence #data