Efficient Convolution Neural Network Software Implementation and DS Preprocessing for AIRBiS

AIRBiS progress meeting†

↑

Background†

With the Coronavirus disease (COVID-19) disrupting large part of the world, the number of patients has grown progressively. For dealing with this serious emergency, accurate diagnosis and fast reporting are two significant mechanisms.

↑

Research Goal†

  1. Search and collect X-ray image datasets containing non-infected (normal) and infected (COVID-19) data given many open sources
  2. Perform pre-processing tasks, e.g. as follows:
    1. denoising
    2. resizing
    3. data augmentation
    4. segmentation
    5. we should also come up with other issues and corresponding strategies
  3. CNN implementation and evaluation
    1. A CNN-based architecture will be adopted for the binary classification task. Once the model learns the feature representation of the input images, a classifier can thus output the diagnosis result, i.e., normal or infected. Framework and language: TensorFlow (Keras), Python
    2. Using the diagnosis (classification) results to evaluate the performance of the CNN (e.g., detection accuracy)
AIRBiS-SW.png
↑

Research Schedule†

DateTask
☑️June 1 - July 15, 2020Making prototype CNN
☑️July 15 - Aug 15, 2020Preprocessing the Dataset
☑️Aug 15 - Aug 31, 2020Argumentation the Dataset
☑️Sep 1 - Oct 31, 2020Evaluation Dataset
☑️Nov 1 - Nov 16, 2020Evaluation CNN
Nov 16 - Nov 25, 2020Writing a paper
Nov 26 - Nov 30, 2020Writing a progress report
Deadline for 1st Draft: 10th(postponed) December 2020International ACM conference
Dec 13 - Dec 23, 2020winter vacation
Dec 24 - Jan 1st, 2021Experiments with different CNN optimizers
7th January, 2021[ACM] Acceptance/Rejection Notification
Jan 7 - Jan 26Preparing a presentation
27-30 January, 2021[ACM] Conference
until Feb 28, 2021research about CNN-SW
Mar or Apr 1, 2021 -Integration of the whole AIRBiS system
Schedule last Updated on: 27/11/2020
↑

Research Progress†

↑

References†

トップ   新規 一覧 検索 最終更新   ヘルプ   最終更新のRSS