machine learning, computer vision and embedded expert
- Verfügbarkeit einsehen
- 2 Referenzen
- 100€/Stunde
- 80999 München
- Weltweit
- it | en | de
- 13.04.2024
Kurzvorstellung
Auszug Referenzen (2)
"A. is a skilled developer and scientist. He integrated very well in the team and made significant contributions to the computer vision system."
1/2019 – 2/2020
TätigkeitsbeschreibungExpert of computer vision. Develop the CV system for retail sector featuring multiple calibrated camera, object detection, human pose estimation, action recognition, 3d reconstruction. Dataset collection and labeling, develop tools for dataset collection and labeling, tools for automated and semi-automated labeling. Optimize execution speed. Interface with cloud sw.
Eingesetzte QualifikationenC++, Docker, Jenkins, Microsoft Azure, OpenCV, Python, PyTorch
"a. was a productive member of the R&D team, contributing IDFs, software packages, co-designed HW/SW prototypes and expert knowledge in HPC/AI."
6/2015 – 3/2018
Tätigkeitsbeschreibung
Expert of numerical algorithms, machine learning, cpu hardware architecture, accelerator hardware architecture.
Guide a team working on high performance numerical software.
Research and design of hw for convolutional neural networks.
Performance modeling, performance estimation. Analysis and benchmarking of prototypes. Individuation of bottlenecks in HW prototypes.
C++, Caffe, Field Programmable Gate Array (FPGA), OpenCV, Python, Verilog HDL
Qualifikationen
Projekt‐ & Berufserfahrung
9/2020 – offen
Tätigkeitsbeschreibung
Manage team of developers, drive product requirement, implement most critical features, collect and cure datasets, train and deploy neural networks, optimize execution speed, port and deploy product on embedded platforms.
The product is a video analytics engine featuring object detection, human pose estimation, tracking, action recognition, anomaly detection, text extraction, etc.
C++, Computer Vision, CUDA, Python
7/2020 – 7/2020
TätigkeitsbeschreibungPort LIDAR object detection DL-based engine from Python to C++ (libtorch) including dependencies. Optimize execution speed, adapt, optimize and retrain network, cooperate with customer on network training.
Eingesetzte QualifikationenC++, CUDA, OpenCV, Python, PyTorch
4/2020 – 8/2020
TätigkeitsbeschreibungDevelop video analytics application, create datasets, train object detector, optimize execution speed, develop algorithms for human action recognition. Develop human pose estimation, integration. Develop code in C++ and Python.
Eingesetzte QualifikationenC++, OpenCV, Python, PyTorch
1/2019 – 2/2020
TätigkeitsbeschreibungExpert of computer vision. Develop the CV system for retail sector featuring multiple calibrated camera, object detection, human pose estimation, action recognition, 3d reconstruction. Dataset collection and labeling, develop tools for dataset collection and labeling, tools for automated and semi-automated labeling. Optimize execution speed. Interface with cloud sw.
Eingesetzte QualifikationenC++, Docker, Jenkins, Microsoft Azure, OpenCV, Python, PyTorch
7/2018 – 11/2018
Tätigkeitsbeschreibung
Main developer. Develop video analytics application, detect and count people, detect people clusters, detect and count crowd, crowd density estimation.
The application used two deep neural networks to detect pepole and estimate crowd density. The application was developed in C++ and used Tensorflow and OpenCV for the CV part.
C++, OpenCV, Python, TensorFlow
6/2015 – 3/2018
Tätigkeitsbeschreibung
Expert of numerical algorithms, machine learning, cpu hardware architecture, accelerator hardware architecture.
Guide a team working on high performance numerical software.
Research and design of hw for convolutional neural networks.
Performance modeling, performance estimation. Analysis and benchmarking of prototypes. Individuation of bottlenecks in HW prototypes.
C++, Caffe, Field Programmable Gate Array (FPGA), OpenCV, Python, Verilog HDL
2/2012 – 5/2015
Tätigkeitsbeschreibung
Work on IBM Blue gene Q prototype, optimize and benchmark massively parallel scientific applications. Research on innovative HW architectures for HPC.
Develop Density Functional Theory (DFT) codes. Research on numerical techniques for DFT. DFT is used to simulate solids and molecules at the quantum level.
C++, Fortran, Linux Entwicklung, Mathematik, Physik, Python
11/2008 – 1/2012
Tätigkeitsbeschreibung
Develop HW and SW for the QPACE project. HW/SW codesign, debug network, develop network Linux device driver, network library with RDMA, develop massively parallel numerical sparse linear solver. Develop innovative solver technique.
Research in numerical solvers, research in Lattice QCD.
C++, Field Programmable Gate Array (FPGA), Fortran, Linux (Kernel), Mathematik, Perl, Physik, VHDL (VHSIC Hardware Description Language)
Zertifikate
Über mich
PhD in Physics, long academic experience in HPC, work on supercomputer prototypes, develop supercomputer HW and SW libraries. Experience in leading technical teams, 15y+ in software development. Languages: C/C++, Python, Fortran, ASM.
In the last 5y i became a computer vision expert and i have contributed computer vision functionality in a number of products for various market segments including: consumer electronics (smartphones), retail, CCTV surveillance and defense.
Weitere Kenntnisse
Persönliche Daten
- Englisch (Fließend)
- Italienisch (Muttersprache)
- Deutsch (Grundkenntnisse)
- Europäische Union
- Schweiz
- Vereinigte Staaten von Amerika
Kontaktdaten
Nur registrierte PREMIUM-Mitglieder von freelance.de können Kontaktdaten einsehen.
Jetzt Mitglied werden