Program Coding Ability
• Language: C/C++/Objective-C, Python/Lua, JAVA/Golang, Assembly, Shell, R, Swift
• Platform: PyTorch/Keras/TensorFlow/MXNet; CUDA; MATLAB/MatConvNet; Docker; ROS; iOS/Android.
• Framework: OpenCV, FFmpeg; TensorRT; SLAM.
Computer Science Ability
• CARSON_WANG AT APPLE COM
• RecursiveTheoryandComputationalComplexity, DiscreteMathematics, ArtificialIntelligence.
• ProbabilityTheory,MathematicalStatistics,RealVariableFunctionTheory;AdvancedAlgebra,MatrixAnalysis;GraphTheory.
APPLE (Shenzhen)
SOFTWARE ENGINEER IN HWTE DEPARTMENT
BiRenTECH (Wangjing)
ALGORITHM ENGINEER IN MODEL QUANTIZATION OF AI-PLATFORM DEPARTMENT
• IindependentlydevelopedthemoduleofmodelquantizationofBireninferenceengine,whichliketheTensorRT. • IindependentlycustomizedYOLOmodelintheDeepStreamonNVIDIAJetsondevice.
InsimationTECH (Jianguomen)
ALGORITHM ENGINEER INTERN IN OBJECT DETECTION GROUP OF TECH DEPARTMENT
Beijing, China
Jun. 2020 - Dec. 2020
• IindependentlycompletedtheimprovementsofoldobjectdetectionmodelbasedonlatestskillsincludingRefinedetandDIoUby the PyTorch and Caffe deep learning frameworks.
Horizon Robotics (Zhongguancun) Beijing, China
ALGORITHM ENGINEER INTERN IN FACE RECOGNITION GROUP OF AIOT DEPARTMENT Jun. 2018 - Nov. 2018
•IindependentlycompletedtheimplementationsofseveralCVPRpapers,like“LDMNet:LowDimensionalManifoldRegularizedNeural Networks” by the MXNet deep learning framework.
Honors & Awards
2016 Second prize, Mathematics Major Group of National College Student Mathematics Competition 2016 Second Prize, National College Students Mathematical Modeling Competition
2016 Second Prize, Anhui University Student Scholarship
2016 Third Prize, Anhui University Student Science and Technology Innovation Award
2013 Second Prize, National High School Mathematics League
I independently completed the improvements of old object detection model based on latest skills including Refinedet and DIoU by the PyTorch
• I independently developed the module of model quantization of inference engine, which like the TensorRT. • I independently customized Y