Skip to content
View pohl-michel's full-sized avatar

Block or report pohl-michel

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
pohl-michel/README.md

Machine learning researcher/engineer with experience in computer vision, time series data analysis and forecasting, and natural language processing. My passion is deep learning. I enjoy using AI to solve society's problems and laying the foundations of next-generation learning algorithms. Concerning the recent AI literature, I am particularly interested in online learning of recurrent neural networks, deep generative modeling, and reinforcement learning. 🧑‍💻 🤖 I have experience in academia and have had the opportunity to explore exciting R&D problems in diverse industries (oil & gas, finance, healthcare, identity & security). Every day, I strive to learn something new by reading research articles or implementing a new idea. I have been living in Japan for more than six years before moving to the UK. 🏯 💂‍♂️ Please do not hesitate to contact me for any matter and let me know if I can help you. I am always happy to connect, talk, and exchange ideas with like-minded tech professionals and computer science and machine learning enthusiasts.

Some of my previous open-source works:

Chest cine MR sequence prediction using PCA and online learning of RNNs Deformable 3D image registration with Lucas-Kanade pyramidal optical flow

Prediction of dynamic sagittal MR cross-sections 6 time steps in advance using sparse 1-step approximation (left: ground-truth, right: prediction).

Calculation of the 3D motion of a lung tumor due to breathing using optical flow.

Time series forecasting with online learning of recurrent neural networks

Prediction of the 3D position of 3 markers placed on the chest 2.1s (7 time steps) in advance using decoupled neural interfaces (the sampling rate is 3.33Hz) to guide the radiation beam during lung radiotherapy.

Pinned Loading

  1. 3D-image-warping-using-Nadaraya-Watson-non-linear-regression 3D-image-warping-using-Nadaraya-Watson-non-linear-regression Public

    Deforming a 3D image according to a given deformation vector field with Nadaraya-Watson regression; 3rd repo in a series of 3 repos associated with the research article "Prediction of the motion of…

    MATLAB 2 1

  2. time-series-forecasting-with-UORO-RTRL-LMS-and-linear-regression time-series-forecasting-with-UORO-RTRL-LMS-and-linear-regression Public

    Prediction of multidimensional time-series data using a recurrent neural network (RNN) trained by real-time recurrent learning (RTRL), unbiased online recurrent optimization (UORO), least mean squa…

    MATLAB 15 5

  3. Lucas-Kanade-pyramidal-optical-flow-for-3D-image-sequences Lucas-Kanade-pyramidal-optical-flow-for-3D-image-sequences Public

    Implementation of the Lucas-Kanade pyramidal optical flow algorithm to register 3D medical images; 1st repo in a series of 3 repos associated with the research article "Prediction of the motion of …

    MATLAB 7

  4. 2D-MR-image-prediction 2D-MR-image-prediction Public

    Future frame prediction in 2D cine-MR images using the PCA respiratory motion model and online learning algorithms for RNNs, and time series forecasting using the latter

    MATLAB 1