eduzhai > Applied Sciences > Engineering >

Classification of diffraction patterns in single particle imaging experiments performed at X-ray free-electron lasers using a convolutional neural network

  • king
  • (0) Download
  • 20210506
  • Save

... pages left unread,continue reading

Document pages: 23 pages

Abstract: Single particle imaging (SPI) is a promising method for native structuredetermination which has undergone a fast progress with the development of X-rayFree-Electron Lasers. Large amounts of data are collected during SPIexperiments, driving the need for automated data analysis. The necessary dataanalysis pipeline has a number of steps including binary object classification(single versus multiple hits). Classification and object detection are areaswhere deep neural networks currently outperform other approaches. In this work,we use the fast object detector networks YOLOv2 and YOLOv3. By exploitingtransfer learning, a moderate amount of data is sufficient for training of theneural network. We demonstrate here that a convolutional neural network (CNN)can be successfully used to classify data from SPI experiments. We compare theresults of classification for the two different networks, with different depthand architecture, by applying them to the same SPI data with different datarepresentation. The best results are obtained for YOLOv2 color images linearscale classification, which shows an accuracy of about 97 with the precisionand recall of about 52 and 61 , respectively, which is in comparison to manualdata classification.

Please select stars to rate!

         

0 comments Sign in to leave a comment.

    Data loading, please wait...
×