Deep Learning Jobs

ReflectiveRAG: Rethinking Adaptivity in Retrieval-Augmented Generation ReflectiveRAG uses a lightweight language-model controller to judge whether the retrieved evidence is sufficient and to reformulate the query when necessary. A contrastive embedding filter then removes redundant or weakly related passages, improving grounding under heavy retrieval noise with limited extra latency. Proposed Project: Adaptive Reflective RAG with a Jury of Judges We already have a working modified RAG system that combines techniques such as hybrid retrieval, reranking, and context compression. These methods improve retrieval accuracy and reduce irrelevant information. However, the current system follows a mostly fixed process and cannot decide for itself when the retrieved evidence is insufficient. The nex...

Sensei AI — Institutional-Grade Stock Analysis Platform I built Sensei AI, a real-time trading intelligence platform covering all 50 Nifty stocks, designed to bring institutional-level AI analysis to retail investors. The system unifies five distinct AI model families into a single scored decision engine (−5.0 to +5.0): - Classical ML — Random Forest with SHAP-based feature attribution for explainable UP/DOWN probability - Deep Learning — Custom LSTM and causal Temporal CNN (PyTorch) for 5-day return forecasting - Reinforcement Learning — PPO agent (Stable-Baselines3) trained via a custom Gymnasium trading environment for BUY/SELL/HOLD actions - Regime Detection — Gaussian Hidden Markov Model to classify BULL/BEAR market states - Financia...

I am looking for an experienced AI/ML developer to build and integrate an advanced AI video transformation feature into my existing portal. The system should create realistic face movements and body motions to simulate live yoga sessions. I will provide a source video, my visual references, and voice/audio samples. The aim is to replace the source face with mine while preserving natural expressions, head movement, body gestures during yoga poses, as well as lip synchronization with the provided voice. It should convincingly mimic a live session's feel. Applicants should have expertise in AI video generation, face replacement, body motion tracking, voice cloning, and backend integration techniques. I'm open to suggestions on using suitable AI models, APIs, or a self-hosted soluti...

Developed a healthcare AI application for lung disease classification using chest X-ray images, combined with an interactive medical chatbot and voice assistant. The classification component analyzes chest X-ray images and predicts the corresponding lung disease category. Alongside the classification model, I was responsible for developing the AI Chatbot and RAG system, allowing users to ask questions about lung diseases, symptoms, and medical information. I implemented Retrieval-Augmented Generation (RAG) to connect the chatbot with a medical knowledge base, enabling it to retrieve relevant information and generate more informative and context-aware responses. The system also included a Voice Assistant to provide a more natural voice-based interaction with the user. Key Features: Che...

Looking for an experienced developer to build an AI-based UAV detection and tracking system for a custom ArduPilot quadcopter using Raspberry Pi 5 and AI HAT+ 2. Scope includes real-time drone detection, object tracking, MAVLink integration, safe autonomous follow/stand-off navigation, telemetry logging, GCS interface, failsafes, SITL testing, and complete source-code handover.

I have a single video frame captured in daylight that shows the front of a vehicle. Unfortunately, the plate is extremely blurry and all normal sharpening tricks have failed. I need a forensic imaging specialist who can push well beyond routine filters—think de-blurring algorithms, AI-based super-resolution, frame averaging, or any other proven technique—to extract a legible plate number. Source material • One “very blurry” video frame pulled directly from the original recording. • Daytime lighting; no night-vision artifacts to worry about. What I would like back • A cleaned, high-confidence still image that clearly shows the license plate characters. • A brief note (a paragraph or two is fine) outlining the process and software you use...

I need a complete solution that pinpoints abnormal behaviour across my wireless sensor network. You will have direct access to three rich data streams—raw sensor readings, detailed network-traffic captures, and accompanying environmental context—so the approach can fuse information rather than relying on a single source. Your job is to design, implement, and validate an anomaly-detection model that flags faulty nodes, suspicious traffic, or out-of-range measurements in near-real time. Feel free to select the technique that best fits the data volume and complexity; I am open to statistical, classic machine-learning, or deep-learning pipelines as long as the final system is transparent and reproducible. Python (NumPy, Pandas, Scikit-Learn, TensorFlow/PyTorch) is preferred for ...

JOB: Computational Biologist / ML Engineer - Brain-Specific Neoantigen Prediction Tool PROJECT OVERVIEW: I am building a brain-specific neoantigen prediction tool for mRNA/LNP/PNA therapeutics. The goal is to predict which mutated peptides (neoantigens) are most likely to be presented by HLA in brain metastasis and trigger an immune response. This is a 7-stage pipeline with a defined step list. WHAT YOU WILL BUILD: Stage 1 - Data Sourcing - Search and download brain metastasis MS immunopeptidomics data from CPTAC, SysteMHC, GEO, PRIDE - Download TCGA primary tumour WES/RNA-seq (20 samples) - Download BrainMetShare brain metastasis WES/RNA-seq (20 samples) - Download GTEx normal brain expression, AFND HLA frequencies, IEDB self-antigens - Record sample metadata (source, cancer type, ance...

I have a collection of 5,000 images that must be annotated with clear, tightly-fitted bounding boxes around the single object of interest in each frame. These labels will feed directly into a new machine-learning pipeline, so consistency and pixel-accurate placement are essential. You are free to work in any mainstream tool such as LabelImg, CVAT, Supervisely or an equivalent—that choice is yours as long as the final export is delivered in a widely-used format (YOLO, COCO JSON or Pascal VOC). I will supply the class list, detailed annotation guidelines and a small set of fully-labeled examples to make expectations crystal-clear before you begin. Deliverables • Complete set of 5,000 bounding-box annotation files in the agreed-upon format • A brief progress log (image c...

I’m building a computer-vision pipeline focused on reliable detection and recognition of human faces and full-body figures. The core of the job is a clean, well-documented Python implementation that can take still images or short video clips and return bounding boxes, class labels, and confidence scores for each detected person. I already have test media and the computing environment; what’s missing is the detection logic itself—ideally leveraging familiar libraries such as OpenCV, TensorFlow, PyTorch, or a proven YOLO/SSD variant. Accuracy on varied lighting and crowded scenes is more important to me than sheer speed, but the code should still run in real time on a modern GPU. Deliverables • Python source code with clear inline comments • Pre-trained weig...

My manuscript on temperature prediction with a Bidirectional LSTM has been returned from a journal with a “major revision” decision. I want to submit a polished, fully compliant revision that answers every reviewer comment and tightens both the science and the prose. The current file is in Word, so all editing should be done with Track Changes enabled. I will supply the reviewers’ reports, original figures, datasets, and any code snippets you may need to replicate or clarify results. Here is what I expect: • A carefully rewritten Word document that integrates the reviewers’ suggestions while preserving my voice and technical accuracy. • Clear explanations—inserted as comments—showing where and how each reviewer point was addressed. &bul...

I want to guide a small group of absolute beginners all the way to the point where they can land an entry-level role in AI, machine learning and generative AI. To get there, I’ll need a mentor who can design and deliver a complete learning journey—covering core theory just enough to ground them, then pivoting quickly into hands-on coding, model training and project work that showcases real-world skills. Here’s what I’m expecting from you: • A clear, modular curriculum that progresses from Python basics through classic machine-learning workflows and on to modern generative models (GPT-style LLMs, diffusion, etc.). • Live or recorded sessions with practical demos, Jupyter notebooks and datasets my trainees can keep experimenting with after class. •...

Project Description: Developed an end-to-end AI-powered medical intelligence platform for analyzing chest X-ray images and assisting with disease prediction. The system uses Deep Learning with DenseNet121 to classify X-ray images into five categories: Normal, Pneumonia, Atelectasis, Cardiomegaly, and Effusion. Integrated Grad-CAM Explainable AI (XAI) to generate visual heatmaps highlighting regions of the X-ray that contributed to the model's prediction. Built REST APIs using FastAPI for image upload, prediction, model information, and health monitoring. The platform also supports prediction history and is designed for integration with AI-assisted medical report generation using LLMs. Key Features 1. Medical Image Analysis – Processes chest X-ray images using deep learning. 2. ...

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...