
Closed
Posted
Paid on delivery
I need an Al-focused application that delivers 100% functionality in analysing and solving geometric patterns and sub patterns and sub to sub pattern at any level in jewellery. The primary objective is to ensure precision and reliable output. The features include: 1. Detecting patterns across different jewellery parts such as any parts of boundaries, textures, cuts, or design elements (including broad patterns, sub-patterns, and intricate sub-to-sub patterns). 2. Identifying low-quality or unclear components within these patterns and seamlessly replacing them with high-quality elements from other sources or jewels. 3. Ensuring replacements naturally blend into the overall piece, maintaining structural integrity and visual coherence. The focus is entirely on achieving accurate functionality with results that meet expectations directly. Experienced Al problem-solving related to geometry or pattern analysis are preferred. Overall Goal of the project: When a customer provides a single image and gives multiple ideas or photos of other parts of the jewels for replacement of the original jewel parts, the iteration and modification process takes days when done in jewellery CAD software. However, these hundreds of experiments and iterations need to be completed in a few seconds, which requires sending very clear final 2d images to jewellery CAD designers. Note: No need to spend any time designing the landing page or logo or anything unnecessary. Just need core skills and full focus to tackle these problem functions and build the right tool. Require building of Deep learning neural framework including the skill of Computer vision. NO Existing AI API buyer build need to from deep learning engineering and computer vision and must familiar with CAD
Project ID: 40765683
15 proposals
Remote project
Active 4 days ago
Set your budget and timeframe
Get paid for your work
Outline your proposal
It's free to sign up and bid on jobs
15 freelancers are bidding on average ₹26,706 INR for this job

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 Python, and similar tools. I have worked with pytorch, and tensorflow to develop DL models, .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
₹35,000 INR in 7 days
7.3
7.3

You need a deep learning system that detects geometric patterns in jewellery images at multiple levels and replaces low-quality sections with high-quality alternatives in seconds, producing CAD-ready output that currently takes days to iterate manually. I would build this from scratch using PyTorch for the neural network framework and OpenCV for computer vision preprocessing. The approach involves training a multi-scale pattern detection model that identifies boundaries, textures and design elements at different hierarchical levels, then a separate network that evaluates component quality and performs context-aware inpainting using features learned from your jewellery dataset. The replacement module would use geometric consistency checks to ensure structural integrity and visual coherence. Output would be high-resolution 2D images optimised for CAD import. Training would require annotated jewellery images showing pattern hierarchies and quality variations, which we would prepare together if you do not have them already. This is custom deep learning engineering work, not API integration. I have delivered AI automation platforms and complex trading systems with real-time decision logic, maintaining a 4.99 out of 5 rating from 150 client reviews on Freelancer across 144 projects, 99.5% on time and 99.8% on budget. I can start immediately and would work through the detection model first, then quality assessment, then the replacement and blending pipeline.
₹37,499.52 INR in 45 days
5.4
5.4

Hi there, I can develop an AI-powered jewellery pattern detector and corrector using Python, OpenCV, and deep learning to identify complex geometric patterns, detect unclear or low-quality sections, and replace them with matching elements from reference images while preserving geometry and visual consistency. I’ll focus on the core computer vision pipeline, accurate image reconstruction, and high-quality 2D outputs for jewellery CAD designers, with clean, documented code and a video walkthrough. Let’s discuss your sample images and expected results so I can suggest the right approach. Best regards, Fahad.
₹12,500 INR in 2 days
4.0
4.0

Hello, The whole point is plain and simple. One jewel photo and few replacement ideas eat days of back and forth in CAD. You want that pile of tries done in few seconds, and a clear 2d image CAD designer can work from. No logo, no landing page. Way I would build it: break the photo into boundary, texture, cut and motif, then into smaller patterns inside those, however deep the design goes. A model trained on your jewels finds the weak bits. Cleaner parts from the other photos, we shape to the same curve and light and blend so the jewel still look like one piece. You get a sharp flat image the CAD person can use. Similar work already shipped. Runner bib detection, https://myraceclip.online. Document defect detection, https://www.freelancer.com/projects/computer-vision/Document-Defect-Detection-System. Micro expression recognition, https://www.freelancer.com/projects/deep-learning/Micro-Expression-Recognition. I wont pretend 100 percent lands on the first try. We check your real samples, note the misses, and tighten until the picture is trustworthy. Own neural nets and vision code, no ready made image API. First detect, replace and export loop in about 3 weeks if photos come early.
₹30,000 INR in 21 days
3.8
3.8

As an AI-focused team with extensive experience in Computer Vision and Machine Learning (ML), we not only design prototypes but deliver reliable production infrastructure, which I believe makes us the perfect fit for your project. We've successfully built similar Deep Learning Neural Frameworks incorporating computer vision for clients across various industries, tackling complex geometric problems and pattern analysis. What distinguishes us is not only the expertise we offer in engineering AI systems, but also the seamless integration of that AI into existing workflows. This is crucial given your need for quick turnaround time in displaying various iterations to your customers. Moreover, our proficiency in hardware exemplifies our ability to handle edge devices—a handy skill when dealing with CAD systems. In conclusion, we don't shy away from challenges that span boundaries and require close attention to detail. With our track record of building agentic AI consistently delivering efficient solutions, coupling with your precise requirement for geometric pattern detection and correction in jewellery designing, I'm confident we can provide a fast, accurate and reliable solution to revolutionize your CAD process. Let's bring our exceptional skills together and create the right tool for you!
₹25,000 INR in 7 days
1.9
1.9

I can deliver 'Geometric deep Pattern detector and correctors AI for Jewellery' end to end — clean architecture, tests, and deployment. I'll start with a short plan, share progress daily, and hand over production-ready code. Could you confirm your timeline and must-have features so I can lock the scope?
₹30,083.33 INR in 7 days
0.0
0.0

Hi, I read your project "Geometric deep Pattern detector and correctors AI for Jewellery" carefully and I'm confident I can deliver it. I have solid experience with Machine Learning (ML), Neural Networks, Image Processing, Computer Vision, Deep Learning, Image Analysis, Diffusion models, CAD / SolidWorks and can complete this within 5 days for INR 12500. Could you share any additional details or examples of what you have in mind? Happy to start right away. Please text me to start a discussion. Best regards
₹12,500 INR in 5 days
0.0
0.0

Thanks for detail. I have worked on custom computer vision models and have solid experience in deep learning and computer vision systems. I understand this workflow well and can get you strong result quickly. I hope to discuss more about project on detail chat. Thanks Marius
₹29,000 INR in 12 days
0.0
0.0

"Sub-to-sub pattern at any level" in jewellery is symmetry: a halo is one prong-and-stone unit repeated round a ring, a filigree band is one motif tiled along a curve, a shank is mirrored. So the detector finds one unit and the rule that repeats it, then recurses: piece, parts, units, primitives. The same rule does the correction: a blurred unit is replaced by the best unit from the piece or your reference photo, placed by the symmetry so scale and perspective match before blending. "100%" comes the only way it does in image work: every change is proposed with a mask, the designer accepts or rejects each in seconds. Everything runs on your machine with open weights. No external AI service. How I'd do it: 1) Name the parts on 50 to 100 of your images; fine-tune an open segmentation model on them, locally. 2) Repeat finder: mirror axes, ring and linear periodicity; find one unit, place all repeats; output a mask tree, one node per level. 3) Quality score per unit (sharpness, completeness, occlusion) to flag what gets replaced. 4) Align the reference part to the target unit from the masks; blend the seam with a locally run inpainting model conditioned on the reference. 5) Handoff: final 2D image, layered masks, change list, measured on held-out pieces you mark. First delivery: all detection levels plus one part type replaced; more types added one at a time. Time: first results on your images within 5 days. I train and deploy my own vision models rather than call APIs: a YOLO system on a live outdoor feed recognising people, animals, damaged walls, wire mesh and garbage, driving alarms and a picking arm; OpenCV pipelines; CAD in KiCad. 15+ years in engineering. Are the input images photos of finished pieces or renders from your CAD software? Renders have clean edges and known lighting, and that changes which model I train first.
₹12,500 INR in 5 days
0.0
0.0

Ramanathapuram, India
Payment method verified
Member since Oct 21, 2019
$2-8 USD / hour
$750-1500 USD
$10-30 USD
₹1500-12500 INR
$250-750 USD
$250-750 AUD
₹12500-37500 INR
₹12500-37500 INR
$750-1500 USD
$15-25 AUD / hour
₹1500-12500 INR
$10-30 USD
$8-15 USD / hour
$15-25 AUD / hour
$2-8 USD / hour
₹1500-12500 INR
$30-250 USD
₹100-400 INR / hour
$250-750 USD
₹12500-37500 INR
$30-250 USD
₹1500-3500 INR / hour
₹1500-12500 INR
$10-30 AUD
₹12500-37500 INR