Day-66 Computer Vision Learning StairNet — Top-Down Semantic Aggregation (Object Detection) by Korea Advanced Institute of Science and Technology Follow me for similar post : 🇮🇳 Ashish Patel Interesting Facts : 🔸 It is published in 2018 WACV, which has already got over 51 citations. 🔸 It outperforms DSSD, SSD, R-FCN, ION, Faster R-CNN & YOLOv2. ------------------------------------------------------------------- 𝗔𝗺𝗮𝘇𝗶𝗻𝗴 𝗥𝗲𝘀𝗲𝗮𝗿𝗰𝗵 : https://lnkd.in/euMerg6 Official Code : https://bit.ly/38imvdw pytorch : https://bit.ly/38jPFZL ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 🔸 One-stage detectors have difficulty in detecting small objects while they are competitive with two-stage methods on large objects since the lower layer that is responsible for small objects lacks strong semantics. 🔸 In this paper, by introducing a feature combining module that spreads out the strong semantics in a top-down manner, the final model StairNet detector unifies the multi-scale representations and semantic distribution effectively. #computervision #arttificialintelligence #innovation