Pembelajaran Mendalam Jobs

Temporal Lesion-Aware Dynamic Gated Multimodal Fusion Framework for DR and DME Analysis Using OLIVES and MMRDR Datasets Framework Overview The proposed framework introduces a Temporal Lesion-Aware Dynamic Gated Multimodal Fusion System for automated analysis of Diabetic Retinopathy (DR) and Diabetic Macular Edema (DME) using multimodal retinal imaging data. The framework combines fundus images, OCT scans, longitudinal retinal information, and optional clinical metadata to improve retinal disease classification, biomarker understanding, and temporal disease progression analysis. Unlike conventional multimodal retinal systems that use static feature fusion, the proposed framework employs a: Dynamic Gated Cross-Modal Fusion Mechanism that adaptively learns the importance of each retinal modal...

Project: ComfyUI / Flux / Z-Image Beauty Filter — Visual Tuning Specialist I'm running a production AI beauty retouching web service (). The full pipeline is built and deployed — what I need is an experienced ComfyUI / Flux / LoRA specialist to dial in the final visual quality. This is a tuning job, not infrastructure work. Pipeline (already running on Modal / A100): Stage 1: Flux 2 Klein 9B + custom Beauty LoRA + ultra_real LoRA Stage 2: Z-Image Turbo + Kook LoRA + ControlNet tile (masked, face/body) MediaPipe FaceMesh eye-preservation blend Custom LAB skin-tone correction (Python) Stage 3: SeedVR2 GGUF upscale to ~2048px Side-by-side composite output Visual problems I need solved: Skin texture too fine / too smooth — natural pore patterns are removed, reads as p...

Complete Lottery Prediction and Betting Automation System (Focused on Loterías y Apuestas del Estado - Spain) 2. System Features 2.1. Historical Data Collection and Update The system must automatically download complete historical results (drawn numbers, draw dates, prize breakdowns by category, accumulated jackpots) from the first draw of each lottery, directly from or reliable associated sources. Specific sources: Euromillones: (since Feb 13, 2004) La Primitiva: (since Oct 17, 1985 – modern version) El Gordo de la Primitiva: (since Oct 31, 1993) Updates automatic at exactly 00:02 the day after each draw, using ethical scraping (BeautifulSoup/Scrapy) with proper user-agent headers to mimic human behavior. Store data in PostgreSQL (structured) or MongoDB (flexible), incl...

Brain disorder detection using Deep Learning Uses MRI brain scan images Detects multiple brain diseases automatically Early disease detection and diagnosis Multi-class classification system Diseases detected: Glioma tumor Pituitary tumor Meningioma tumor Alzheimer’s disease Healthy brain Uses Convolutional Neural Networks (CNN) Uses Transfer Learning techniques Models used: CNN VGG16 ResNet50 Dataset collected from Kaggle MRI datasets Combined brain tumor and Alzheimer datasets 1000 images used for training 250 images used for testing Image preprocessing and normalization Pixel values converted from 0–255 to 0–1 Feature extraction using convolution layers Pooling layers reduce computation Fully connected layers perform classification Softmax activation used for multi-clas...

1. Smart CT Scan Analyzer An AI system that analyzes lung CT scan images and detects possible lung cancer signs early. Main Features Upload CT scan or X-ray images AI detects lung nodules Cancer risk prediction Heatmap highlighting suspicious areas Doctor dashboard Patient history tracking PDF medical report generation Technologies Frontend: Vue.js / React Backend: Node.js or Python Flask AI Model: TensorFlow / PyTorch Database: MySQL / PostgreSQL Image Processing: OpenCV AI Models CNN (Convolutional Neural Network) ResNet50 YOLO for object detection Users Doctors Radiologists Hospitals Patients 2. AI + IoT Lung Monitoring System A smart healthcare platform connected to wearable devices. Features Real-time breathing monitoring Oxygen level tracking AI predicts lung disease risk Emergen...

We are looking for a hands-on Senior Computer Vision Architect to scale our live multi-camera video analytics platform from 100 to 1,000+ cameras across NVIDIA DeepStream and distributed edge device architectures. The ideal candidate must have strong production experience with NVIDIA DeepStream SDK, TensorRT optimization, edge AI deployments, and large-scale camera infrastructure. Responsibilities: • Design and manage a scalable 1,000-camera architecture with high availability and low latency • Optimize DeepStream pipelines (NvInfer, NvTracker, NVDEC, GStreamer) for stable FPS under heavy load • Deploy an manage CV models on Jetson or equivalent edge devices • Fine-tune YOLO models and export optimized TensorRT engines • Build and improve infrastructure using Doc...

I am building a research-grade pipeline that focuses first and foremost on Diabetic Retinopathy (DR) grading, while keeping an eye on downstream extensions such as DME detection, retinal biomarker discovery, longitudinal modelling, cross-dataset generalisation and explainable retinal AI. Data at hand • Thousands of colour fundus photographs sourced from OLIVES, MMRDR and my own curated set. • Structured patient medical history accompanying a subset of those images. Core objectives 1. Train and validate a model that assigns the correct severity level to every DR image. 2. Surface early-stage indicators that could alert clinicians before conventional thresholds are reached. 3. Provide per-patient progression tracking so that sequential visits can be plotted and foreca...

I have a growing library of MP4 recordings of casino-grade slot machines and I need a reliable way to turn each session into structured data. For every spin I want the script to capture the start time, end time, bet size, win amount, any bonus triggers, and jackpots, then tally an overall spin count. The videos cover many different games, so the screen layout shifts from title to title; your solution therefore has to locate each region of interest on its own instead of relying on hard-coded coordinates. Because bonuses and jackpots are marked with visually distinct graphics, you can leverage that cue when classifying events. Accuracy matters more than raw speed, but I still expect processing to run unattended once configured. Please build a computer-vision / OCR pipeline that: • Au...

I am putting together a small educational AI/ML project that detects diabetic retinopathy on the publicly-available OLIVES dataset. The goal is not only to build a working multimodal model (fundus images plus any supporting clinical metadata you find useful) but also to showcase clear explainability and rigorous evaluation so the project can be presented in an academic setting. Scope of work • Prepare the OLIVES dataset, handle any class imbalance, and document the preprocessing pipeline. • Design and train a multimodal architecture of your choice in Python—PyTorch, TensorFlow or another modern framework is fine—as long as the code is clean and reproducible. • Produce the quantitative metrics I need: Accuracy, Precision, F1-score, AUC and Cohen Kappa on a ...

I aim to build an advanced, research-grade Virtual Try-On (VTON) engine that can realistically place Tops (e.g., shirts, blouses), Bottoms (e.g., pants, skirts), and Full outfits (e.g., dresses, suits) onto a human model from nothing more than a single 2-D photograph. The workflow should centre on state-of-the-art deep generative techniques—diffusion models, flow-matching, and transformer-based architectures—so the final renders look genuinely photo-realistic, preserve garment texture, and respect body pose and occlusion. The system will be trained on a curated dataset I already possess, then fine-tuned to accept JPEG and PNG uploads at inference time. Clean, modular PyTorch (or equivalent) code, a reproducible training pipeline, and inference scripts that run on a single high...

I have a collection of general-purpose images and need a complete Python-based pipeline that extracts meaningful features and classifies each image accurately. The project centres on image feature extraction and subsequent classification, so solid experience with OpenCV, scikit-learn or a deep-learning stack such as TensorFlow or PyTorch is essential. You will begin by deciding on (and justifying) an appropriate feature strategy—traditional descriptors like SIFT/ORB, transfer-learning from a CNN, or another proven method—then train and validate a classifier that reaches reliable accuracy on a held-out test set. Clean, well-commented code and clear, reproducible training steps are critical because I need to retrain the model as new data arrives. Deliverables • Python sour...

The project centers on building a production-ready TensorFlow 2.x model that classifies tabular data delivered to us through an internal API. I have the API specifications and sample payloads ready; you will turn those streams into a clean training pipeline, engineer the right features, and iterate until the classifier meets our performance targets in real-world tests. Scope of work • Data pipeline – pull the API data, handle preprocessing, and produce TensorFlow-friendly datasets for train/val/test splits. • Model development – design, train, and tune a deep learning architecture suitable for tabular inputs (e.g., wide & deep, Transformer, or other proven structures). • Optimization – experiment with hyperparameters, regularization, and callback...