David Santiago Morales Norato

David Santiago Morales Norato

About

Detail

Deep learning engineer | Junior Research Scientist | Computational imaging
Bucaramanga, Santander, Colombia

Contact David regarding: 
work
Full-time jobs
Starting at USD700/month
connect_without_contact
Finding mentors
Finding co-founders
groups
Networking

Timeline


work
Job
school
Education
folder
Project
auto_stories
Publication

Résumé


Jobs verified_user 0% verified
  • S
    Scientific Committee Member
    STSIVA 2024
    Feb 2024 - Jul 2024 (6 months)
    • Participated in the scientific committee member of STSIVA 2024, showcasing significant research in signal processing, spanning theoretical foundations to practical applications in data science. • Organized keynote speaker sessions, elevating the conference's academic and professional influence. • Conducted thorough peer reviews of submitted papers, upholding high standards of quality and relevance.
  • P
    Co-founder
    PyColibri
    Aug 2023 - Current (2 years 11 months)
    • Co-founder and leader of PyColibri, an open-source PyTorch library for computational imaging (CI) tasks, mainly leading the team of six engineers to build and maintain the optics and mathematical models behind PyColibri. • Tools Used: Python, PyTorch, Git, GitHub Actions, Sphinx
  • Universidad Industrial de Santander
    Junior Imaging Scientist
    Universidad Industrial de Santander
    Jan 2023 - Current (3 years 6 months)
    • Designed two novel regularization terms in PyTorch, based on the condition number of the sensing matrix and the modulation transfer function of the point spread function, for end-to-end learning in diffractive optics, enhancing model stability and performance. • Implemented a multimodal architecture in Pytorch based on Vision and Language Transformers (CLIP) for Human Action Recognition. • Developed in PyTorch a reconstruction algorithm based on Plug-and-Play ADMM optimization frame work, with a Denoising Convolutional Neural Network prior for solving inverse problems in imaging, such as spectral video (4D) reconstructions from compressive measurements (2D). • Developed an implicit neural representation architecture in PyTorch t
  • Universidad Industrial de Santander
    Machine Learning Engineer
    Universidad Industrial de Santander
    Sep 2022 - Dec 2022 (4 months)
    • Designed and implemented a deep learning model integrating Convolutional Neural Networks and Long Short-Term Memory architectures using TensorFlow to predict organic carbon content, achieving up to a 10% improvement in estimation accuracy with data preprocessing and feature engineering in scikit-learn.
  • Tecnalia Colombia
    Data Scientist
    Tecnalia Colombia
    Oct 2021 - Oct 2022 (1 year 1 month)
    • Developed advanced analytical models and simulations to optimize the design of a hyperspectral camera payload for the FACSAT-3 mission, a project led by the Colombian Air Force in collaboration with leading research institutions. • Achieved a 50% reduction in payload size by applying optical system optimization techniques and leverag ing ray tracing simulation software, ensuring compatibility with the constraints of a 1.5U CubeSat while preserving high spatial and spectral resolution. • Authored a peer-reviewed journal article documenting the design methodology, optimization process, and system characterization and performance over segmentation models implemented in Tensorflow and Scikit learn.
  • C
    Machine Learning Engineer
    Chill & Chat
Education verified_user 0% verified
  • Universidad Industrial de Santander
    Master of Science in Applied Mathematics
    Universidad Industrial de Santander
    Feb 2023 - Current (3 years 5 months)
  • Universidad Industrial de Santander
    Bachelor of Science in Systems Engineering
    Universidad Industrial de Santander
    Jan 2017 - Jan 2021 (4 years 1 month)
Projects (professional or personal) verified_user 0% verified
  • Universidad Industrial de Santander
    End-to-end design of DOEs for privacy-preserving for multimodal action recognition.
    Universidad Industrial de Santander
    Jan 2023 - Jan 2025 (2 years 1 month)
Publications verified_user 0% verified
  • S
    Shift-variant PSF characterization of doe-based imaging systems via implicit neural representations,
    Aug 2023
  • C
    Compressive Spectral Video via Wavelength-Dynamic Coded Aperture
    Aug 2023
  • Universidad Industrial de Santander
    Multi-domain discrete spectral image fusion via implicit neural representation
    Universidad Industrial de Santander
    Aug 2023
  • Universidad Industrial de Santander
    Hyperspectral camera as a compact payload architecture for remote sensing applications
    Universidad Industrial de Santander
    Jan 2023
  • L
    Learning spectral initialization for phase retrieval via deep neural networks
    Jan 2022 - Jan 2025 (3 years 1 month)
  • D
    Deep Phase Retrieval by a Learnable Filtered Spectral Initialization
    Aug 2021