SIVAKUMAR VELLINGIRI

Having 10+ years of extensive experience in Image Processing, Computer Vision, Machine Learning and Deep Learning with edge platform integration.

EXPERTISE:

• USECASES: Object detection, Object tracking, Lane detection, Face recognition, License plate recognition, Driver Monitoring System, Pose estimation, Speech-to-text, GEN_AI (Threat detection using vision language model), Distance Estimation, Collision Warning, Traffic Sign board Detection, Intrusion Detection, Tripwire Detection, People counting, People Loitering, Motion detection, Smart motion detection, Abandoned & Missing Detection, Heat map for crowd density, Crowd detection, Queue Monitoring System, Video Tampering , Scene change detection.
• FRAMEWORKS: Tensorflow, Pytorch, Darknet, Caffe, Keras
• ADDITIONAL EXPERTICES: OpenCV, Triton, Deepstream, TensorRT, ONNX, TAO
• ARCHITECTURES:

DETECTION & CLASSIFIERS: ssd_mobilenet(v1,v2,v3), Yolo(v1,v2,v3,v4,v5-"n,m,s,x,l",v6,v7,v8,nas), Yolo-tiny(v2,v3), Faster RCNN, Facenet, FAN, Dlib, BlazeFace, MTCNN, Arcface, Retinaface, Custom Models

CLASSIFIERS: Mobilenet(v1,v2,v3), Inception(v1,v2,v3), Resnet(18,34,50,101,152,164,1202)

AUDIO PROCESSING: Whisper model(tiny, base, small, medium, large)

• PLATFORMS: Qualcomm-qcs(410, 605, 610), Nvidia-Jetson(Xavier NX, TX2 NX, Nano, Orin), Ambarella-CV25, CV28, CV52, Sigma Star-336Q, 338Q, Intel-Myriad-X, Xilinx-K26B, Alveo.