๐ฅ 6x Linkedln Top Voice | AI Research Scientist & Chief Data Scientist at IBM | Generative AI Expert | Author - Hands-on Time Series Analytics with Python | IBM Quantum ML Certified | 11+ Years in AI | MLOps | IIMA |
๐๐ฎ๐-๐ญ๐ญ๐ฑ Computer Vision Learning ๐๐ฒ๐ฒ๐ฝ ๐ฆ๐ป๐ฎ๐ธ๐ฒ for Real-Time Instance Segmentation by Zhejiang University Follow me for similar post : ๐ฎ๐ณ Ashish Patel Interesting Facts : ๐ธ This is a paper in CVPR2020 with over 36 citations. ๐ธ Outperform MNC, FCIS, ORCNN, Mask R-CNN, PANet ------------------------------------------------------------------- ๐๐บ๐ฎ๐๐ถ๐ป๐ด ๐ฅ๐ฒ๐๐ฒ๐ฎ๐ฟ๐ฐ๐ต : https://lnkd.in/eKi7bAx code : https://lnkd.in/eApEqFb ------------------------------------------------------------------- ๐๐ ๐ฃ๐ข๐ฅ๐ง๐๐ก๐๐ ๐ธ This paper introduces a novel contour-based approach named deep snake for real-time instance segmentation. ๐ธThe basic idea of deep snake. Given an initial contour, image features are extracted at each vertex (a). Since the contour is a cycle graph, circular convolution is applied for feature learning on the contour (b). The blue, yellow and green nodes denote the input features, the kernel of circular convolution, and the output features, respectively. Finally, offsets are regressed at each vertex to deform the contour to the object boundary (c). #computervision #artificialintelligence #deeplearning
๐ฅ 6x Linkedln Top Voice | AI Research Scientist & Chief Data Scientist at IBM | Generative AI Expert | Author - Hands-on Time Series Analytics with Python | IBM Quantum ML Certified | 11+ Years in AI | MLOps | IIMA |
3yFor all previous posts visit: https://github.com/ashishpatel26/365-Days-Computer-Vision-Learning-Linkedin-Post