Youssef Aidani Email & Phone Number

Berlin, De

Video Super Resolution Benchmarking: Built a configurable and asynchronous benchmarking pipeline forboth deep learning based and hardware accelerated VSR algorithms.◦ Video Processing & Codecs: Developed video processing data-gens that perform video i/o ,pixel formatconversion, codec operations for both raw and compression video data for production and benchmark purposes.◦ Video Super Resolution Technologies: Integrated super resolution technologies (IntelSR, MaxineSR,WebexSR) for both video production and performance evaluation.◦ Video Quality Assessment: Developed modules to asses video quality in both video and frame level.◦ Video Data Preprocessing: Developed different cli modules to prepare datasets for super resolution modeltraining using different denoising, deblurring, artifact removal background removal techniques.◦ Face Normalization: Built video caller face normalization module to control per frame bit rate and call requiredbandwidth.◦ Video Super Resolution Models: Currently working on training, optimizing and testing different video andimage deep learning architectures for different video restoration problems.

https://www.upwork.com/freelancers/youssefaidani

Istanbul, Turkey

Worked on on-shelf product management using computing vision techniques: The techniques used were Image Classification Using different architectures of CNN, Product Detection Using Yolo V3 and V4. The adressed problems: - Product Out Of Stock detection- On shelf prodcut Presence- Planogram Compliance Detection

Ankara, Turkey

Using RNN's to make energy consumption forecasting.Given a huge dataset of smart meters readings associated with their time stamp, I worked on time Series Modeling of household energy consumption using different LSTM and GRU architectues. Techniques applied were: - Hyperparameter Tuning- Regularization- Model Validation and Selection- Data Visualizatin

Ankara

Parallel computing techniques to solve high rank linear systems