Qualitative approach to the task.
The project fully covered all the objectives.
The code is clearly written and well documented.
A pleasant impression of joint work.
I will continue to consult with the following projects.
W = 3000 (width) H = 2000 (height)
V = 5 (number of variants - blurred images)
List of RGB images (W x H): [..., IMG_%IN%_%SM%.jpg, ...]
%IN% - index number (001, 002, ... )
%SN% - serial number (SN == 00 is origin, 00 < SN <= 0%V%)
By default: N=40 M=20
Input: [N x N x 3]
Output: [M x M x 3]
M <= N
splitImage(pathToModel:str, imgPath:str, saveDir:str)saveDirapplyToImage(pathToModel:str, imgPath:str, savePath:str)savePath:strТочность не сильно критична (70% и выше достаточно), главное что бы всё работало.
Budget: 4000 UAH Deadline: 5 days
1) Cut
2) Augmentation by rotation [0°, 90°, 180°, 270°]
3) Preparing a data set
4) Create the NS
5) Teaching
6) Write the cutting algorithm, processing with NS and assembly into the entire image
Основная задача: написать нейронку (любой топологии) на Tensorflow которая принимает на вход размытoe изображениe размером NxN, и выдаёт откорректированное изображениe MxM, с точностью от 70% ,где N < 60 пикселей, M <= N. Для примера взять N=40пх, M=20пх.
Входные данные: набор больших изображений (3000x2000), где одно оригинальное, и пять к нему с разной степенью размытости.
Под задачи:
- Подготовить данные для тренировки - написать функцию разбития больших изображений на мелкие тайлы (NxN) для тренировки на них.
- Две функции: splitImage и applyToImage для применения обученной модели.
Датасета надо сформировать, на основе больших изображений (1 оригинал и 5 размытий), в итоге, да получатся мелкие изображения одно - заблуренное рамером NxN, и одно не заблуренное MxM. Тоесть надо написать функцию генерации таких мелких пар имея на входе одно большое изображение не размытое и несколько больших размытых. Условно получить по 1000 мелких с одного сета.
We are looking for a specialist or a team to implement AI solutions in the communication of the CT and MRI medical center. Requirements: integrate AI into telephony for handling incoming and cold calls; ensure natural Ukrainian-speaking communication with minimal pauses; integrate an AI assistant into the Binotel chat or propose an effective alternative; automate responses to inquiries, initial consultations, and scheduling for examinations. Important: the medical field requires accurate information gathering before CT or MRI, particularly regarding the examination area, preparation, referrals, and possible contraindications. The solution must be empathetic, professional, and allow for the transfer of complex cases to an operator. In your response, please send relevant cases, a description of the proposed solution, and estimated timelines. Experience in medical projects will be an advantage.
It is necessary to develop an application in MATLAB that can process images/videos, identify individual objects, analyze their characteristics, and, if necessary, use machine learning methods to automate the analysis. Desirable: confident knowledge of MATLAB; experience in Computer Vision / Image Processing; experience with Machine Learning;
Developer for creating AI bots - multi-agent system and automations in AI application. What is needed: proficiency in n8n; experience integrating OpenAI and other AI models via API; working with webhooks, REST API, and Supabase; ability to create reliable AI workflows and automations. Create multi-agent systems from scratch or with a framework AI Agents, MCP, RAG
Counting of finished products and peopleNon-conforming products. Based on a high-performance line, ensure accurate counting of throughput, workforce participation, identification of non-conformities, and context of stops/reconfigurations, providing ready analytical information.
Good day, what is needed: 1. Conduct an analysis of the processes (personal of the manager) and the business. 2. Consider where (in which processes) AI assistants can be implemented. Where it will be appropriate and effective. 3. Develop a plan. We are engaged in the publication of scientific articles.