Unlimited Email and Phone number
Everything you need to reach more inboxes and close more deals
Unlimited Contact Only at$79
Mehmet D.
Lead Data Scientist @CDM Smith
Signup · Get unlimited contacts
WORK HISTORY
Lead Data Scientist @CDM Smith
Boston, MA, US
Computer Vision-based Intelligent Traffic Analysis for FHWA Vehicle Classification • Led development of an end-to-end Computer Vision pipeline for vehicle detection and axle classification based on FHWA Vehicle Classification standards • Utilized detection models, tracking algorithms, and multiple classification models • Ensured scalability and optimized performance for GPU-enabled AWS Virtual MachinesAdvance Traffic Behavioral Analysis • Engineered a Computer Vision algorithm to compute vehicular speed and detect wrong-way driving at the lane level using video input. • Utilized pavement markings and frame analysis within predefined regions of interest for accurate detection. • Developed algorithms to calculate vehicle speed based on frame count and distance traveled. • Implemented a data processing pipeline to generate lane-level sensitive data for comprehensive analysis. • Leveraged Computer Vision techniques for efficient processing and accurate results.
EDUCATION
University of Massachusetts Lowell
Master of Science (MS), Computational and Applied Mathematics
University of Massachusetts Dartmouth
Master of Science, Data Science
Yıldız Teknik Üniversitesi
Bachelor of Engineering (B.Eng.), Mathematical Engineering
SKILLS
ABOUT MEHMET D.
As a Lead Data Scientist with a robust background in AI, Deep Learning Modeling, and Computer Vision Engineering, I specialize in leading and developing cutting-edge end-to-end AI/ML/Data solutions. Actively involved in project exploring LLM models, particularly in Generative AI, and langchain technology.I am deeply passionate about data science. With 6+ years of experience, I specialize in computer vision projects across diverse industries, spanning transportation, sewer, habitat, environment, and traffic safety. Proficient in integrating innovative concepts, I excel in creating high-performing models that consistently surpass expectations and deliver tangible results. Experienced in deploying AI to edge devices through edge computing to optimize deep learning algorithms and enhance performance across diverse applications.My experience with text analytics, entity extraction, sentiment analysis, data engineering, data visualization, and NLP on the Azure Cognitive platform has played a significant role in my exploration of emerging technologies like LLM and generative AI. Model optimization for real-time edge deployment Using Nvidia APIs for real-time performance: Deepstream (GStreamer based, build custom plugins), TensorRT. Vehicle detection, classification and tracking Abnormal traffic behavior analysis: lane change counting, wrong-way detection, congestion detection Speed estimation and vehicle occupancy detection Sending all metadata to the cloud using Kafka/MQTT brokers App and models deployment at the cloud/edge Model training on the cloud (AWS/Azure) using transfer learning Experience on low-cost AI edge devices: Nvidia Jetson, Intel Neural Compute Stick, Google Edge TPU. Using small/fast CNN models for real-time detection and classification like Yolov3/4/5/6, SSD Mobilenet, SSD Inception, Resnet. Skills- Programming Languages: Python (Tensorflow, Numpy, Pandas, Sklearn, Matplotlib, Seaborn), R Data: Microsoft SQL Server, PostgreSQL, MongoDB Data Visualization tools: Qlik, Tableau, PowerBI Practical Understanding of Deep Learning algorithms/ techniques, Neural Networks, Natural Language Processing, Hyperparameter tuning, Tree-based Algorithms (Decision Tree, Random Forest), Clustering techniques (kNN) Cloud and Orchestration: Amazon EC2, S3, Docker, Docker Compose, Microsoft Azure Frameworks : PyTorch, TensorFlow, Keras, OpenCV, CUDA
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.