Ashish Patel 🇮🇳’s Post

𝗗𝗮𝘆-𝟯𝟲𝟮 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻 𝗟𝗲𝗮𝗿𝗻𝗶𝗻𝗴 𝗦𝘄𝗶𝘀𝘀 𝗙𝗲𝗱𝗲𝗿𝗮𝗹 𝗜𝗻𝘀𝘁𝗶𝘁𝘂𝘁𝗲 𝗼𝗳 𝗧𝗲𝗰𝗵𝗻𝗼𝗹𝗼𝗴𝘆 𝗟𝗮𝘂𝘀𝗮𝗻𝗻𝗲 𝗵𝗮𝘀 𝗽𝘂𝗯𝗹𝗶𝘀𝗵𝗲𝗱 𝗣𝗥𝗜𝗠𝗘: 𝗔 𝗙𝗲𝘄 𝗣𝗿𝗶𝗺𝗶𝘁𝗶𝘃𝗲𝘀 𝗖𝗮𝗻 𝗕𝗼𝗼𝘀𝘁 𝗥𝗼𝗯𝘂𝘀𝘁𝗻𝗲𝘀𝘀 𝘁𝗼 𝗖𝗼𝗺𝗺𝗼𝗻 𝗖𝗼𝗿𝗿𝘂𝗽𝘁𝗶𝗼𝗻𝘀(𝗗𝗮𝘁𝗮 𝗔𝘂𝗴𝗺𝗲𝗻𝘁𝗮𝘁𝗶𝗼𝗻 𝗧𝗲𝗰𝗵𝗻𝗶𝗾𝘂𝗲) Follow me for a similar post: 🇮🇳 Ashish Patel ------------------------------------------------------------------- 𝗜𝗻𝘁𝗲𝗿𝗲𝘀𝘁𝗶𝗻𝗴 𝗙𝗮𝗰𝘁𝘀 : 🔸 Paper: PRIME: A Few Primitives Can Boost Robustness to Common Corruptions 🔸 This paper is published arxiv2021. 🔸 The data augmentation method introduced in the paper: "PRIME: A Few Primitives Can Boost Robustness to Common Corruptions". PRIME is a generic, plug-n-play data augmentation scheme that consists of simple families of max-entropy image transformations for conferring robustness to common corruptions. PRIME leads to significant improvements in corruption robustness on multiple benchmarks. ------------------------------------------------------------------- 𝗜𝗠𝗣𝗢𝗥𝗧𝗔𝗡𝗖𝗘 🔸 Despite their impressive performance on image classification tasks, deep networks have a hard time generalizing to many common corruptions of their data.  🔸 To fix this vulnerability, prior works have mostly focused on increasing the complexity of their training pipelines, combining multiple methods, in the name of diversity.  🔸 However, in this work, we take a step back and follow a principled approach to achieve robustness to common corruptions. We propose PRIME, a general data augmentation scheme that consists of simple families of max-entropy image transformations.  🔸 We show that PRIME outperforms the prior art for corruption robustness, while its simplicity and plug-and-play nature enables it to be combined with other methods to further boost their robustness.  🔸 Furthermore, we analyze PRIME to shed light on the importance of the mixing strategy on synthesizing corrupted images, and to reveal the robustness-accuracy trade-offs arising in the context of common corruptions.  🔸 Finally, we show that the computational efficiency of our method allows it to be easily used in both on-line and off-line data augmentation schemes. #computervision #artificialintelligence #data