C# or Python AI Frame Detection Engine

@Microlent

Hi, I have gone through the C# or Python AI Frame Detection Engine requirements. It looks straightforward at first, but frame selection criteria, model accuracy, processing speed, and API integration can cause issues later. I can outline a practical approach using computer vision and deep learning in Python or C#, with FastAPI if an API layer is required. I have over 10 years of experience in Python, C# programming, machine learning, neural networks, computer vision, deep learning, and API development, and I'm excited to be part of your project. Let's discuss your frame detection requirements, input formats, expected outputs, and performance targets in the chat so I can recommend the most practical and cost-effective solution. ~ Rajesh

@sajjadtaghvaeifr

Hello, I trust you're doing well. I am well experienced in machine learning algorithms, with nearly a decade of hands-on practice. My expertise lies in developing various artificial intelligence algorithms, including the one you require, using Matlab, Python, and similar tools. I hold a doctorate from Tohoku University and have a number of publications in the same subject. My portfolio, which showcases my past work, is available for your review. Your project piqued my interest, and I would be delighted to be part of it. Let's connect to discuss in detail. Warm regards. please check my portfolio link: https://www.freelancer.com/u/sajjadtaghvaeifr

@seefattechnologi

Hi, This requires a real-time computer vision pipeline rather than simple object detection. I will build a scalable inference engine that processes video frames, detects sport type, logos, players, jerseys, scoreboards, ball possession, and returns structured results through a FastAPI or ASP.NET Core API with multi-worker support. My Approach: • Develop optimized CV models for logo, player, jersey, scoreboard, and ball detection with GPU/CPU compatibility. • Build a scalable worker-based API with in-memory state, continuous frame processing, and secure deployment. • Deliver complete source code, demo dashboard, documentation, and Windows-ready deployment. One question: Will the training datasets for players, logos, and jerseys be fully labeled, or is annotation also part of the project? Please ping me to get started and deliver outstanding results. Thanks!!!

@ShaikhAneesa

Hi there, I understand you need a real-time computer vision engine that continuously processes sports video frames, identifies sports, teams, players, logos, kits, scores, and ball possession, then returns structured results through a scalable API. I am confident I can deliver a modular, production-ready solution that is accurate, secure, and designed for continuous processing across both GPU and CPU environments. My approach will be to build a multi-worker inference pipeline using FastAPI with independent workers capable of processing concurrent video streams. The engine will combine YOLO/OpenCV for object detection, OCR for scoreboard extraction, logo and kit classification models, player number recognition, jersey-to-player mapping using your supplied roster files, and ball possession tracking. Each worker will maintain its own session state, return structured JSON responses in real time, and scale horizontally as demand grows. The solution will include configurable model loading, logging, error recovery, Windows compatibility, REST APIs, and a lightweight demo dashboard for live validation. Will the training datasets for logos, jerseys, and players be fully provided, or should the engine also support incremental model training as new teams and seasons are added? I'm ready to start immediately. Warm Regards, Aneesa.

@corpmember29

Hello, I HAVE WORKED ON SIMILAR AI COMPUTER VISION, OBJECT DETECTION, SPORTS ANALYTICS, AND VIDEO PROCESSING SYSTEMS AND CAN SHARE RELEVANT EXAMPLES. I completely understand your requirement for a scalable AI frame detection engine capable of analyzing continuous video frames and identifying sports, team logos, player names and jersey numbers, home/away kits, scoreboard information, and ball possession in real time. I can build a modular, multi-worker solution using Python (FastAPI) or C# (ASP.NET Core) with GPU/CPU support, optimized inference, secure APIs, in-memory processing, and a demonstration dashboard for validation. The architecture will be designed for easy scaling while maintaining high accuracy, performance, and reliability on both Windows and Linux environments. I have 10+ years of experience in Python, C#, Computer Vision, OpenCV, YOLO, TensorFlow, PyTorch, FastAPI, ASP.NET Core, OCR, and AI model deployment. I have developed high-performance AI systems for object detection, image recognition, real-time video analysis, and scalable backend services with clean, maintainable code. I WILL PROVIDE 2 YEARS OF FREE ONGOING SUPPORT, COMPLETE SOURCE CODE, WE WILL FOLLOW AGILE METHODOLOGY, AND I WILL ASSIST YOU FROM INITIAL PLANNING TO SUCCESSFUL DEPLOYMENT. I am available to start immediately and look forward to discussing your project. I eagerly await your positive response. Thanks, Christina

@maxscender

I can help you build this as a modular C# backend that avoids heavy GPU dependency for core tasks—logo detection and jersey recognition can run on pre-trained ONNX models, while sport classification and scorebar OCR use lightweight classifiers that work CPU-side. The multi-worker architecture fits naturally into .NET's TPL Dataflow or Channels, letting you spin up N workers that each hold their own model instances and state, re-initiable via a simple API endpoint. For player identification, I’d parse your provided text file into a lookup structure and match detected jersey numbers against it frame-by-frame, handling partial occlusions with a confidence threshold fallback. The frontend can be a minimal Blazor or plain HTML/JS dashboard that visualizes returned JSON—sport type, logo list, player tags, home/away, score overlay text, and ball possession—updating per frame batch. Windows-compatible, no external GPU requirement unless you choose to optionally accelerate. Source code structured cleanly so your team can integrate it without untangling dependencies.

@Muhammadzeesha59

Hello Sir/MAM I am a Skilled Full Stack Developer. Having rich experience in Java , C++ , C , C# , Python , Eclipse , Sql , Mysql , .Net ,Oracle , Object Oriented Programming , Data Structure , Algorithms, Linux , Windows , Cloud , Azure , Ubuntu , OpenAI , Desktop Applications. Web Development I have a perfect grip on “Artificial Intelligence” “Automation” , and work in “Machine Learning” Deep Learning “Computer Vision ” Object Detection”. My track record as demonstrated in my 100% job completion and 5-star review rating showcases My ability to deliver exceptional results on time and with utmost quality I believe that my skill set makes me the ideal candidate for this project Please come on chat we will discuss more about this I will be waiting for your reply . Thanks and Best Regards

@nandasunny449

Dear Hiring Manager, I am excited to apply for your project. With 8.6+ years of experience in .NET Full Stack development, I have successfully built and maintained enterprise web applications, RESTful APIs, and scalable business solutions for clients across various industries. My expertise includes: Languages: C# Frameworks: ASP.NET Web Forms, ASP.NET MVC, .NET Core, WinForms, WPF, AngularJS, Angular 18 Frontend: React, JavaScript, jQuery, HTML5, CSS3, Bootstrap, AJAX, Figma to HTML/CSS APIs: REST APIs, Web API, WCF, Windows Services, Third-Party APIs, Facebook API Databases: Microsoft SQL Server, MySQL Payment Integrations: Stripe, WePay Tools: Telerik, DevExpress, Git, Bitbucket, Azure DevOps, SourceTree Testing: Selenium I am a reliable developer with strong communication skills and a proactive approach to problem-solving. Whether you need a new application, enhancements to an existing system, API integrations, or ongoing maintenance, I am confident I can deliver results that exceed your expectations. I am available to start immediately and would welcome the opportunity to discuss your project in more detail. Thank you for considering my application. I look forward to hearing from you. Best Regards, Sunny

@topcreator100

Hello, As an experienced developer skilled in Python and other relevant technologies for the sport-related project such as yours, I am confident in my ability to deliver an elegant, high-performance solution. I have a demonstrated background in building scalable web applications and APIs, which will be put to good use as we construction your backend engine for analyzing sport frames. The fact that I possess a deep understanding of database design, performance tuning, and scalability is also beneficial to this project, as it factors in our requirement to work with unlimited workers. Moreover, my proficiency in working across various platforms including both GPU- and non-GPU Windows/Linux environments ensures that your deliverable won't be limited by technology or performance constraints. This goes hand-in-hand with the on-time delivery approach I prioritize; I make certain to allocate adequate time for thorough testing and quality checks to ensure secure, reliable applications with no loopholes - a significant desire of yours as highlighted in the project description. In terms of intellectual property ownership, rest assured Sors Markets Ltd will retain full rights to the source code and all related IP. Finally, I deem it essential for you to know that I always give my portfolio projects (and all their components) my absolute best - both in terms of technical expertise and dedication to client satisfaction. We've got a solid chance Thanks!

@toriquldev123

Hello Dear, I’m Md Toriqul Islam, and I’m excited to partner with you. I can dive into your project immediately. With over 10 years of experience in Python, AI, computer vision, API development, and scalable backend systems. I understand you need a secure sports recognition engine that processes video frames, detects sports, teams, logos, players, scores, and ball possession through scalable workers, returning real-time results via FastAPI or ASP.NET Core. I have rich experience in Python, FastAPI, OpenCV, YOLO, OCR, REST APIs, multithreading, GPU optimization, and Windows deployment. I’m ready to start immediately and would be happy to discuss your architecture and timeline. Looking forward to hearing from you. Best regards, Md Toriqul Islam

@trm66614

I am excited about the opportunity to develop your C# or Python AI Frame Detection Engine. Understanding the project’s intricacies, I will create a robust system to efficiently process and identify various sports frames and related data. Using FastAPI or ASP.NET Core, I will ensure the engine reliably detects types of sports, logos, players, and match info, while effectively managing multiple workers. The design will prioritize both GPU and non-GPU environments, guaranteeing seamless performance across Windows systems. To ensure it meets your high standards, I will implement rigorous testing for security and reliability, avoiding any bugs or loopholes. I understand the importance of the source code and will ensure it is well-structured for your integration. My experience in similar projects ensures I can deliver on these requirements and more. For your review, I can provide my portfolio showcasing relevant work. Could you specify any specific sports or logos you want prioritized for detection? Looking forward to the possibility of collaborating with you on this exciting project. Best, Talha

@mayankinnovative

The challenge here isn't object detection alone—it's building a real-time vision pipeline that can process continuous video frames, combine multiple recognition tasks, and return structured sports metadata with low latency while remaining stable across long-running sessions. I'd approach this as a modular inference engine with independent detectors for sport classification, logos, player number/name recognition, home/away kit detection, scoreboard extraction, and ball possession. A FastAPI-based worker architecture would allow multiple workers to process incoming frames concurrently, maintain session state in memory, and return aggregated results through a consistent API. The training pipeline would be organized around your folder structure, making it straightforward to extend teams, logos, and sports without changing the core engine. The solution would be validated on Windows with and without GPU acceleration, accompanied by a lightweight dashboard demonstrating continuous frame processing, worker management, and live results. I've worked on computer vision systems involving object detection, OCR, image classification, and API-driven inference pipelines where performance, reliability, and maintainability were central to the implementation. I can also walk through the architecture and relevant portfolio examples during the interview.

@Aaleen110

The core of this job is accurately identifying and extracting specific visual cues from video frames, which means building a Computer Vision pipeline. I'd use Python with OpenCV for image processing tasks and feed those processed frames into a Deep Learning model trained for object detection and classification, likely a YOLO or Mask R-CNN variant, to pick out the logos and player numbers. For player identification from shirt numbers, once a number is detected, I'd use a mapping of numbers to players for each sport, so the model would need to output both the number and its bounding box. I'll assume the sports data you provide will be in a common video format like MP4 or AVI and that the API for feeding frames will be a simple RESTful POST request. Identifying home/away kits from player wears requires a classifier that can distinguish subtle color differences and patterns associated with specific team kits, so that would be a separate classification model. Extracting match info from the top left corner would involve OCR, so I'd integrate Tesseract OCR, and then parse the output for team names and scores. Determining who is holding the ball requires motion tracking and analysis of player-ball proximity, so I'd employ optical flow algorithms to track ball movement and player positions. I am a Preferred Freelancer on Freelancer with a 5.0 rating, 100% on time and 100% on budget. What is the expected resolution of the screen video frames? Let's schedule a short call on Freelancer to confirm the video frame resolution and discuss the specific API endpoint structure.

@a4logic

Hello, With over a decade of experience in C# programming, my team and I at A4Logic offer you limitless possibilities in executing this challenging project. Our extensive track record of building software solutions across various domains - including complex API integrations and large-scale data management - positions us exceptionally well to address your distinct needs. We understand the importance of an AI engine that is robust, adaptable, and able to cope with high demand - all while ensuring clean, bug-free code. Moreover, our proficiency extends beyond C#, as we also have substantial expertise with Python libraries emphasizing AI & deep learning frameworks. Whether It's FastAPI for enhanced performance and flexibility or ASP.NET Core API for seamless Windows integration, we're confident about delivering a solution tailored specifically for the project. Lastly, we recognize the sensitivity associated with secure data management and uphold rigorous standards to guarantee its protection. Our robust testing procedures will ensure that your program is reliable and impenetrable. Thanks!

@rosscarter1011

hello, Of the six detection tasks you listed, five are established computer vision work, but one is genuinely hard, and it is worth being honest about which. Sport type, logos, and home-versus-away kit are all reliable image classification and detection. Reading the score overlay is OCR. But player identification from shirt numbers during live play is the real challenge, because numbers get blurred by motion, hidden by other players, folded by the body, and only visible from behind, so that one needs the most training data and realistic accuracy expectations, not a promise of perfection. So I would build this as a pipeline of specialised models rather than one model doing everything, each stage detecting what it is good at, with the detections fed forward. Your training folder structure fits this cleanly, since each sport, logo, and kit set trains its own detector. FastAPI wrapping the engine handles the multi-worker frame delivery and scaling you described, running with or without GPU. The "who is holding the ball" piece I would treat as ball detection plus nearest-player logic, which is achievable but accuracy depends heavily on camera angle. Roughly how much labelled training data exists per team for the kits and logos, since that sets realistic accuracy? And what frame rate must each worker keep up with, since real-time versus near-real-time changes the model choices? have a nice day.

@emilesd617

Hi, I am a full stack AI developer with 8 years of rich experience in software development, with a background in . I am familiar with Python, C# Programming, Computer Vision, Machine Learning, Deep Learning, Neural Networks, FastAPI, API Development, etc. For this project, I will build a scalable frame detection engine with reusable workers, real time API processing, GPU and CPU support, and accurate detection of sports, teams, logos, players, scores, and ball possession with a simple demo interface. I'm an individual freelancer and can work on any time zone you want. Please contact me with the best time for you to have a quick chat. Looking forward to discussing more details. Thanks. Emile.

@ImranAli01

Hi There, I got that you are looking for a computer-vision engine that processes continuous sports frames, identifies the sport, logos, players and shirt numbers, home/away kits, match scores, and the player holding the ball, while supporting multiple reusable workers. This is what I can help you with, let's chat. My approach is to build the backend in Python with FastAPI, using OpenCV and YOLO-based detection models with OCR for shirt numbers and scoreboards. I’ll structure the training pipeline around your SportsDetection folders and text-based player data, then create worker processes that can be reinitialized and scaled from 3 workers upward. The engine will support GPU and CPU Windows environments, maintain frame-level state, and return structured sports data through the API. As final deliverables you will receive the complete source code, FastAPI backend, computer-vision detection pipeline, training configuration, reusable worker architecture, API integration, simple demonstration front end, Windows-compatible deployment setup, and documentation. One thing I'd like to confirm before we start: will the supplied sports data include enough labeled examples for player, logo, kit, and ball detection, or should I also prepare the training/annotation pipeline? I’d be happy to discuss the architecture and demonstrate a working approach on Freelancer. Cheers, Imran

@starttechno

This project needs a practical, reusable vision pipeline rather than a one-off frame classifier. I can build it in Python with FastAPI, using modular detectors for sport type, club logos, shirt numbers/player names, home/away kits, score overlays, and ball possession. The training loader will read your SportsDetection\Football\2026\Ball, Logos, and Wear folders, while the supplied player text files will populate the number-to-name mapping. I’ll design the API around reusable worker sessions: each worker receives continuous frames, maintains the latest identified state in memory, and returns structured JSON to the frontend. Multiple workers will be supported from the start, with configurable concurrency and clean reinitialization. The implementation will support Windows with or without CUDA, selecting GPU inference when available and CPU inference otherwise. I’ll also add validation, authentication-ready endpoints, request limits, logging, error handling, and safe file/model loading so the engine is suitable for integration into a larger system. The delivery includes source code, setup instructions, sample training data handling, API documentation, and a simple browser dashboard showing live detections and worker status. I have 10+ years of experience, 249+ delivered projects, and extensive work with Python, FastAPI, . Muhammad Saad

@alizain721

Hello, I will implement a FastAPI based frame detection engine that accepts continuous frames via API, runs reinitiable workers (start at 3), and returns structured lists: sport type, detected logos, identified players with numbers, home/away wear, top-left match overlay text, and current ball holder. My bid of 135 GBP matches your stated budget. I previously built a FastAPI sports pipeline that ran 3 concurrent workers for live football streams and delivered a Windows-compatible demo and source code. Deliverables will include source code that runs on Windows, model training from your folder structure, a simple front end demo, worker management (scalable beyond 3), basic hardening and unit tests, and support for GPU and CPU environments. If you share a short sample training folder and a few frames I will deliver a working demo on your Windows machine with full source within 48 hours, no commitment. Happy to jump on a quick chat. Regards Ali Zain