Muhammad Salman T.

I develop computer vision, deep learning, and AI automation solutions for real-world applications across sports, healthcare, finance, security, and industrial sectors. My core expertise includes YOLO-based object detection and tracking, video analytics and scoring systems, OCR and document intelligence, and AI chatbots for business workflows. From analyzing sports footage and tracking objects in video streams to digitizing ECG records and automating document-heavy processes, I build Python-based AI systems that transform complex data into meaningful, structured outputs. 𝗪𝗵𝗮𝘁 𝗜 𝗦𝗽𝗲𝗰𝗶𝗮𝗹𝗶𝘇𝗲 𝗜𝗻 𝗖𝗼𝗺𝗽𝘂𝘁𝗲𝗿 𝗩𝗶𝘀𝗶𝗼𝗻 & 𝗢𝗯𝗷𝗲𝗰𝘁 𝗗𝗲𝘁𝗲𝗰𝘁𝗶𝗼𝗻 ▸ Object detection, classification, and multi-object tracking using YOLO and deep learning. ▸ Sports analytics, player tracking, performance analysis, and video-based scoring systems. ▸ Vehicle detection, traffic monitoring, and automatic number plate recognition (ANPR). ▸ Industrial defect detection and automated visual quality inspection. 𝗩𝗶𝗱𝗲𝗼 𝗔𝗻𝗮𝗹𝘆𝘁𝗶𝗰𝘀 & 𝗥𝗲𝗮𝗹-𝗧𝗶𝗺𝗲 𝗩𝗶𝘀𝗶𝗼𝗻 ▸ CCTV analytics, crowd detection, and activity recognition. ▸ Multi-camera video analysis and object tracking pipelines. ▸ RTSP stream processing and real-time inference. ▸ Custom vision models for monitoring, measurement, and event detection. 𝗢𝗖𝗥 & 𝗗𝗼𝗰𝘂𝗺𝗲𝗻𝘁 𝗔𝗜 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻 ▸ Automated information extraction from financial, insurance, legal, and medical documents. ▸ Invoice, receipt, and purchase order processing. ▸ Layout-aware document parsing, table extraction, and scanned document digitization. ▸ Structured data extraction with validation, confidence scoring, and human-review workflows. 𝗔𝗜 𝗖𝗵𝗮𝘁𝗯𝗼𝘁𝘀 & 𝗪𝗼𝗿𝗸𝗳𝗹𝗼𝘄 𝗔𝘂𝘁𝗼𝗺𝗮𝘁𝗶𝗼𝗻 ▸ AI chatbots for finance-related services and customer interactions. ▸ Python-based automation for repetitive, data-intensive business processes. ▸ AI-powered information retrieval and processing for business workflows. ▸ Integration of AI capabilities into existing applications and operational systems. 𝗛𝗲𝗮𝗹𝘁𝗵𝗰𝗮𝗿𝗲 𝗔𝗜 & 𝗦𝗶𝗴𝗻𝗮𝗹 𝗣𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴 ▸ ECG digitization using computer vision, deep learning, and digital signal processing. ▸ Image and signal analysis for healthcare data processing. ▸ Classification, pattern recognition, and anomaly detection. ▸ AI pipelines for converting complex healthcare data into structured digital formats. 𝗔𝗜 𝗗𝗮𝘁𝗮 𝗔𝗻𝗻𝗼𝘁𝗮𝘁𝗶𝗼𝗻 & 𝗠𝗼𝗱𝗲𝗹 𝗧𝗿𝗮𝗶𝗻𝗶𝗻𝗴 ▸ Image and video annotation for supervised learning. ▸ Bounding boxes, polygons, and instance or semantic segmentation. ▸ Pose estimation, keypoint annotation, and tracking datasets. ▸ Dataset preparation, quality validation, and model training pipelines. 𝗧𝗲𝗰𝗵 𝗦𝘁𝗮𝗰𝗸 ▸ Programming & Processing: Python, OpenCV, digital signal processing. ▸ Computer Vision: YOLO, Faster R-CNN, DETR, DeepSORT, ByteTrack, MediaPipe. ▸ Deep Learning: PyTorch, TensorFlow, transformer-based models. ▸ OCR & Document AI: PaddleOCR, Tesseract, TrOCR, LayoutLM, Azure Document Intelligence, AWS Textract. ▸ Annotation Tools: CVAT, Label Studio, Roboflow, Supervisely. ▸ Deployment: FastAPI, Docker, ONNX, TensorRT, AWS, Azure, GCP. 𝗜𝗻𝗱𝘂𝘀𝘁𝗿𝗶𝗲𝘀 𝗜 𝗪𝗼𝗿𝗸 𝗪𝗶𝘁𝗵 ▸ Sports & Performance: Player tracking, video analysis, and scoring systems. ▸ Healthcare: ECG digitization, medical data processing, and image analysis. ▸ Finance & Insurance: AI chatbots, document extraction, and workflow automation. ▸ Security & Transportation: Video surveillance, object tracking, and vehicle recognition. ▸ Manufacturing & Industrial Operations: Visual inspection, defect detection, and quality control. 𝗪𝗵𝗮𝘁 𝗬𝗼𝘂 𝗖𝗮𝗻 𝗘𝘅𝗽𝗲𝗰𝘁 ▸ Clear communication and practical technical recommendations. ▸ Solutions designed around your data, use case, and performance requirements. ▸ Maintainable Python code and well-structured AI pipelines. ▸ Realistic scoping, testing, and attention to real-world data quality. Whether you're solving a complex computer vision problem, automating data-intensive processes, or building a custom AI system, I focus on delivering solutions that work beyond the prototype. Share your requirements, and let's identify the right technical approach for your goals.