Data augmentation jobs
...desert, coastline, mountain range, or any other natural scene we decide on—and return the correct label with strong accuracy. Here’s what I need from you: • A well-structured dataset or clear guidance on sourcing and curating one (public sets are fine as long as licensing is respected). • A training workflow in Python using a mainstream framework such as TensorFlow or PyTorch, complete with data-augmentation, fine-tuning, and validation steps. • Trained model weights plus inference code that runs on CPU or GPU with a single command. • A concise README explaining environment setup, training parameters, and how to add new classes later. • Evaluation metrics (precision, recall, confusion matrix) so I can judge performance quickly....
...is no less than 95 % on voices recorded in typical everyday settings, so the model must stay robust whether the user is in a bustling office, walking along a busy street, or standing in a crowded venue. Here’s what I need from you: • An end-to-end speaker-recognition pipeline—data preprocessing, feature extraction, model training, and inference—implemented in Python with a mainstream deep-learning stack such as PyTorch or TensorFlow. • Noise-handling techniques (e.g., spectral subtraction, data augmentation with synthetic noise) integrated so the final model meets the 95 % identification benchmark on a held-out, noisy test set that we will agree on. • A concise report explaining architecture choices, training parameters, and ...
...signal-to-noise ratio drops. Python is preferred for the pipeline, and frameworks like PyTorch or TensorFlow are perfectly acceptable. Please work with publicly licensable datasets or clearly state any proprietary material you intend to use, and describe your noise-augmentation strategy up front so I can vet it. Deliverables • A trained speaker-recognition model capable of handling noisy audio • A lightweight API or CLI demo that accepts a wav/mp3 file and returns the speaker ID plus a confidence score • A short technical report covering data sources, preprocessing, training procedure, and validation results, including accuracy on a withheld noisy test set (≥90 % target) Share your proposed workflow, expected timeline, and any questions you need...
...**Physics-based synthetic waveform generation** using lung-mechanics equations to generate realistic Paw, flow and volume signals, initially focusing on normal breathing and double triggering. 2. **Synthetic dataset creation** with waveform signals, images, labels and physiological parameters. 3. Addition of realistic **noise, artefacts and physiological variability**, followed by diffusion-based augmentation. 4. Development of AI models using **time-series and waveform images**, including Transformers, CNNs, Bi-LSTM and Swin Transformer. 5. **Multimodal fusion** of signal and image features for PVA classification. 6. **Synthetic-to-real domain adaptation** using DANN, CORAL and MMD. 7. **Explainable AI (XAI)** using Grad-CAM, Integrated Gradients, Score-CAM and Attention Rollout...
...category. My end goal is straightforward: a reliable model that can classify every new photo I add, with clear metrics that show its accuracy. I haven’t fixed the exact number of categories yet; we can review the sample set together and decide what makes sense. You are free to work in Python with libraries such as TensorFlow, Keras or PyTorch, and any supportive tools for preprocessing, augmentation, and evaluation. I’ll share the image folders via a cloud link once we start. Please hand over: • Clean, well-commented source code (notebook or script) • The trained model weights or a reproducible training script • A short report summarising accuracy, confusion matrix, and key insights If everything runs smoothly on my machine and the results...
...folder of labelled pictures into a production-ready image-classification model. This project sits squarely in machine learning: the main task is to build, train, and validate a solution that can correctly place each incoming image into its predefined category with solid, measurable accuracy. Here is how I see the work unfolding: • Prepare and clean the image dataset, applying sensible augmentation so the network generalises well. • Select or design a CNN architecture (transfer-learning with ResNet, EfficientNet, or a custom model—whichever you believe will perform best) and implement it in Python using TensorFlow or PyTorch. • Train, tune hyper-parameters, and track performance with clear metrics; I want the training notebook or script fully reprodu...
...qualified leads for B2B services and managing advertising budgets efficiently. **Key Responsibilities* * Plan, launch, and manage LinkedIn Ads and Google Ads campaigns. * Target U.S. companies and decision-makers such as founders, CEOs, CTOs, CIOs, and engineering leaders. * Build campaigns for services including: * Custom software development * Software development outsourcing * Staff augmentation and dedicated engineering teams * AI and automation solutions * MedTech and healthcare software * QA, DevOps, and engineering services * Conduct keyword and audience research. * Manage targeting, negative keywords, bidding, retargeting, and budget allocation. * Test audiences, advertisements, offers, and landing pages. * Coordinate with the marketing and design teams on ...
I have a single LinkedIn post that needs to be placed in front of the exact people who buy the kind of work we do at Xponexus Engineering. My only goal is lead generation, with a secondary aim of bringing those same people to our website for a closer look at our remote structural engineering support, CAD/BIM services and resource-augmentation offer. Here is the audience I must reach: • UK-based structural or civil engineering consultancies, architectural practices with in-house structural teams, design consultancies and small-to-mid multidisciplinary firms, typically 2 – 250 staff. • Decision-makers only: Managing Directors, Engineering Directors, Technical Directors, Structural Engineering Managers, Principal Structural Engineers and BIM/CAD Managers. • No stu...
...(Normal, Arrhythmia, or Myocardial Infarction) together with a confidence score. The workflow should be wrapped in a clean web interface—Flask or Streamlit are both acceptable, so choose whichever lets you move fastest without compromising clarity. What I expect: • A well-trained deep-learning model, full source code, and a clearly commented training notebook that shows every preprocessing, augmentation, and tuning step. • The curated dataset (or links plus usage licence) you used, with an explanation of its size, class balance, and any cleaning you performed. • The web app code that loads the saved weights, accepts image uploads, runs inference, and displays the prediction with the probability. • A concise installation guide (Linux/Windows), pl...
...archaeological problem. The mission is clear: build an AI model capable of reading an undeciphered script, pushing the limits of CNNs and related deep-learning techniques. Your day-to-day focus will be on designing, training, and iterating the model—from data pipelines in Python through to hyper-parameter tuning in TensorFlow or PyTorch—until we reach publishable accuracy. For context, you’ll be working with: • Images of ancient scripts • Transcriptions of ancient texts • Annotated historical data Expect hands-on experimentation with OCR, image augmentation, and transfer learning, plus frequent reviews with archaeologists to validate outputs. I’ll provide mentoring, a negotiable monthly stipend, and a completion certif...
I’m commissioning an in-depth, critical article on S.B. Divya’s novel “Machinehood.” The goal is to move well beyond a plot recap; I want a sharply argued, text-supported analysis that unpacks the book’s core ideas—its commentary on artificial intelligence, gig-economy labour, bio-augmentation, and the ethics of personhood—and shows how those themes are delivered through Divya’s narrative choices, pacing, and point-of-view shifts. To keep the piece useful for editors and curious genre readers alike, ground every point in concrete examples: quote or reference specific scenes (please include page numbers from the U.S. print edition where possible) and, where it genuinely deepens the discussion, draw brief comparisons with works such as ...
...with an external AWS engineering partner; * the prospect has agreed to a real business conversation, not only accepted a calendar invite. What we do not want * scraped, unverified lead databases; * mass messaging without personalization; * fake or incentivized appointments; * meetings with junior employees or irrelevant roles; * companies looking only for individual contractors; * pure staff augmentation opportunities; * payment based only on calendar bookings without qualification....
I need an end-to-end web application that takes a single chest X-ray image and instantly classifies it as Normal or Pneumonia. You will build, train, and deploy the full pipeline: • Model development – Train on the public Kaggle Chest X-Ray Images (Pneumonia) set (≈5,800 images) using a modern CNN architecture in Python. – Handle data cleaning, augmentation, train/validation/test splits, and hyper-parameter tuning to achieve competitive accuracy. – Save the final model and include concise documentation of the training process for reproducibility. • Backend API – Expose the trained model through a lightweight REST or gRPC service (Flask, FastAPI, Django, or similar) that accepts a single image upload and returns the pre...
...drift off-target too often. I need your help to push raw and benchmarked accuracy noticeably higher without sacrificing latency. You’ll dive into the existing Python codebase (TensorFlow and a light PyTorch utility are already in play), audit the current model pipeline, then propose and implement improvements—be that better tokenisation, a more suitable transformer architecture, advanced data augmentation, or smarter post-processing. The interface overlays results on a visual HUD, so clean, deterministic outputs matter; hallucinations or fluffed confidence scores show up instantly to users. Deliverables: • Refactored or newly trained model files ready for production • Updated inference script compatible with the HUD’s API endpoints &bull...
I’m building a proof-of-concept that lets an ultra-low-power board—specifically an ESP32—capture image data and immediately flag early-stage oral cancer. Because the hardware budget is tiny, I need a novel, highly compressed AI architecture that still reaches clinically useful accuracy and inference speed at the edge. The device will ingest live image data from a camera module, run on-device preprocessing, then execute an AI-based model (think pattern-recognition and other machine-learning techniques rather than simple thresholding) fast enough to give instant feedback without off-loading to the cloud. Latency, memory footprint, and power draw must all stay within the ESP32’s limits. If you’re interested, send a detailed project proposal ou...
...perfectly acceptable so long as the final classifier is accurate and quick to infer on standard CPU hardware. Key deliverables: • Clean, commented source code for training and inference • The trained model weights (or exported SavedModel / .pt file) • A short README explaining environment setup, how to retrain with new data, and a one-line CLI or REST example for running predictions • Basic metrics report: training/validation accuracy, confusion matrix, and any augmentation techniques applied Acceptance criteria: • ≥90 % overall validation accuracy across all categories • Clear, reproducible results when I run the provided script on a fresh machine If you have questions about the dataset format or need additional samples, let me ...
...and bread-crate detection, hovering below 50 % accuracy. With only ~100 training images (each holding 30–40 crates), I need to push performance past 94 % without relying on power-hungry cloud instances. I’m open to every practical angle—tighter algorithmic tuning, smart preprocessing and creative data augmentation—so long as the final solution can run locally on a mid-range GPU or even CPU if possible. Feel free to experiment with lighter YOLO variants, pruning, quantisation, mosaic augmentation, rotation/flip tricks, colour tweaks or any other ideas you trust; I care about the end result and the ability to reproduce it on my hardware. Acceptance criteria • Provide the updated model weights, training script and a concise README. &b...
...Services Custom Software Development Web Application Development Mobile App Development (Android & iOS) SaaS Product Development CRM Development ERP Solutions AI & Machine Learning Solutions Automation Solutions Product Management Consulting UI/UX Design QA & Software Testing Cloud Solutions & DevOps Salesforce Development & Consulting LMS & EdTech Solutions Staffing & Resource Augmentation Digital Marketing & Growth Solutions What We Are Looking For We are looking for partners who can bring: Software Development Projects Mobile App Development Projects SaaS Product Development Projects Enterprise Digital Transformation Projects CRM/ERP Implementation Projects AI & Automation Projects Dedicated Development Team Requ...
I’m building an end-to-end flower–image recognition system that runs as a web application. Users will be able to upload photos from their local drives, and the backend will instantly identify the flower species with high accuracy. Here is the scope I need covered: • Data pipeline – curate or expand an existing public flower dataset, handle cleaning, augmentation, and train/validation/test splits. • Model development – implement a Convolutional Neural Network in Python (TensorFlow / Keras or PyTorch) tuned specifically for multi-class flower classification. I’m open to transfer-learning from ImageNet-based architectures if that speeds convergence. • Evaluation – provide precision, recall, F1, and confusion matrix on ...
...must be written manually with a corporate tone. No AI text and no AI images are allowed. You must also agree not to disclose that this work was done by you or your agency. --- Content Requirements Write complete content for: Company introduction Mission and vision Values Leadership message Why choose us IT services overview AI and automation Cloud and DevOps Data and analytics Cybersecurity Staff augmentation Industries served Engagement models Work process Legal and compliance note Contact information Content must be original, professional, and enterprise-grade. --- Design Requirements Corporate, modern layout Infographics and clean visuals No AI images 10 to 20 pages Editable file required Final PDF required Design similar to Wipro, Infos...
...engineer who can architect, train, and iterate on deep-learning models that perform both medical-imaging analysis and diagnostic support. The scope covers X-ray, MRI, and CT data, so you should be comfortable handling multimodal image pipelines and the differing pre-processing each modality demands. You will start from a clean slate: selecting or designing network architectures in Python, building them with PyTorch or TensorFlow, and setting up a repeatable training environment that lets us experiment rapidly. Once a strong baseline is in place, I want to see steady, research-driven improvements—new loss functions, data-augmentation ideas, self-supervised techniques, or anything that reliably drives accuracy upward while keeping the models clinically robust. ...
Our development team in Ahmedabad is moving into the next round of builds and I need clean, reliable power supply systems engineered from the ground up. The work ranges from choosing the right topology for each application through to producing fully documented schematics and layout drawings. Intermediate-level Microsoft Visio skills are essential because every diagram, layer and cross-reference will be handed straight to production and compliance for approval. You will interpret our electrical requirements, size components, run efficiency calculations, and capture every decision inside Visio so nothing is lost between design and assembly. Expect to collaborate with our in-house engineers by video call and chat; strong, clear communication will keep iterations fast and productive. Deliver...
...the enlarged, pulse-pumping heart needs to stretch, squash, and beat in sync with his emotions just like the Jim Carrey scene, only adapted to my footage. I’m aiming for a mixed style: grounded enough to sit naturally on the live-action plate, yet cartoonish enough to keep the over-the-top charm. You are free to combine classic compositing tools (After Effects, Nuke, Houdini, Blender) with AI augmentation if that speeds things up or sharpens the look—just keep it seamless. Key deliverables • Final graded video in 6K (6144 × 3456) ProRes or equivalent high-bit-rate codec • Optional preview MP4 (watermarked) for review rounds • Project files or layered final comp so I can archive or tweak later Acceptance criteria • Heartbeat visible a...
...that would take classical supercomputers millennia to solve. 3. Biotechnology and Neurotechnology The convergence of biology and technology (often called "BioTech") is creating revolutionary tools: CRISPR gene editing advancements and synthetic biology. Brain-computer interfaces (BCI), such as Neuralink’s work on direct neural communication. Personalized medicine powered by AI analysis of genomic data. 4. Sustainable and Green Technology Climate change has driven massive research investment into: Next-generation batteries and energy storage. Advanced solar photovoltaics and perovskite cells. Carbon capture and utilization technologies. Fusion energy research (e.g., ITER project and private companies like Commonwealth Fusion Systems). 5. Advanced Materials and ...
...If energy charge and capacity charge tariffs are adjusted → revenue and returns update automatically The model should support interchangeable scenario analysis without breaking formulas. 5. Scenario & Sensitivity Analysis Please include: * Base / Best / Worst case scenarios * Sensitivity tables for: * CAPEX * Tariff * Interest rate * Energy yield * Capacity factor * Battery augmentation/degradation * Energy charge tariff * Capacity charge tariff * Optional tornado chart/dashboard is a plus 6. Deliverables * Excel financial model (.xlsx) * Clear instructions or notes tab explaining: * Inputs * Outputs * Formula logic * How to run scenarios * Sample filled-in case study for demonstration Preferred Experience: * Renewable energy financial...
...it as a native-first alternative that feels natural to C# developers yet mirrors the familiar TensorFlow/Keras workflow. Scope of functionality The first milestone centres on the same pillars that make Keras intuitive: • Neural Network Building – an object model that lets me compose layers, define sequential or functional graphs, and introspect parameters. • Data Preprocessing – utilities for batching, shuffling, augmentation and dataset pipelines comparable to tf.data. • Model Evaluation – training loops, metric tracking, checkpointing and model save/load logic. Runtime expectations The library will be used in standalone fashion; it only needs to expose a public API that any C# solution can reference. Low-level compute may targ...
I have an existing home that needs to be documented in FreeCAD and a planned version of the same home that I would like drawn up in the same file. The level of detail I’m after sits comfortably between a simple outline and full-blown architectural drawings: walls, openings, roof shape and key dimensions accurate enough for later augmentation, but nothing structural or interior-design heavy at this stage. For the existing house I need floor plan, that if needed can be transformed into elevation views, But I have basic elevation that may be OK; the planned version can stay floor-plan only for now. Interior layouts, finishes and furniture are not required—But need wet room item: washer, dryer, shower bath. loo etc. Deliverables • FreeCAD (.FCStd) file containing two ...
...Para 4: XAI + summary of our proposed multimodal work with 3 contributions 2. SECTION III. METHODOLOGY [Edit ONLY these subheadings] 4.1 System Architecture: Pipeline text + describe Fig 1: Data → Preprocessing → ResNet-50 + ClinicalBERT → Fusion → Classifier → Grad-CAM 4.2 Dataset Description: Create Table 1 like reference - Source: MIMIC-CXR, Pairs: 1,495, Classes: 14, Split: patient-level 70/10/20. Add sentence: "Due to computational constraints, we used a curated subset following standard practice [ref]." 4.3 Data Preprocessing: Bullets - Resizing 224×224, Normalization ImageNet, Augmentation, Balancing 4.4 Mathematical model: Add equations for Softmax + Weighted BCE Loss for multi-label. Remove "causal&...
Augmentations of the shadow diagrams CAD of the Ocean Street Project
I am building a Windows-based computer-vision system that pulls images and related text from PDF and XML sources, then trains a deep-learning model to recognise those images with high accuracy. Everything happens in Python, so the entire training and evaluation pipeline—including data loaders, augmentation, model definition, loss functions, metrics, and experiment tracking—must be written in clean, modular Python code that runs smoothly on a local workstation or GPU server. You will connect preprocessing routines to the raw documents, structure datasets for efficient loading, and iterate through model architectures (CNNs, transformers or other modern backbones). Precise reporting of training curves, confusion matrices, and F1/accuracy scores is essential; the pro...
I’m looking for an experienced Vendor Manager who can take full ownership of our IT-focused staff-augmentation partnerships. Day to day, you will be the primary point of contact for each supplier, scheduling and leading the recurring check-ins that keep everyone aligned on head-count, contract terms, and upcoming skill needs. A large part of the role is welcoming new vendors. I’ll rely on you to steer the entire onboarding process—from the first introductory call through credential verification, tool access, and final approval—so new partners can start submitting qualified candidates without delay. Because all of our engagements revolve around technology talent, prior experience dealing with IT or tech service providers is essential. You should already ...
I want a crisp, visually polished capability statement that I can send to prospective U-S clients when introducing our IT staffing firm. The document must tell a clear, confident story in a single glance, so please bring both strong copy-editing skills and sharp layout instincts. Content to include • Company overview • Core competencies in IT staffing and project augmentation • Past performance highlights (I will supply bullet points) • Proof of our DFW MSDC minority certification and our active SAM registration Branding Our color palette and typography are already live on ; mirror that look so the statement feels like a natural extension of our existing material. If you need exact hex values I can fetch them, but an eyedropper from the site should ...
We are a company specializing in Staff Augmentation and Business Process Outsourcing (BPO) services for the insurance, financial, and commercial sectors in the United States. The company operates delivery centers in LATAM and also develop technology solutions focused on contact center automation and system digitalization. We are seeking a digital marketing agency or a specialist with proven B2B demand generation experience to support the promotion of our Staff Augmentation services and manage our digital channels. The ideal agency should be fully bilingual, with all content developed in English and targeted to the U.S. market. Required capabilities include: * Google Ads and performance marketing expertise * Conversion tracking through pixels, events, and funnels * Funnel a...
...that spots vehicles with high accuracy. Because I do not yet have a labeled dataset, the job begins with sourcing or capturing varied vehicle images and annotating them in classic YOLO format. Once the dataset is in place, the next step is to train and validate the network—YOLOv5 or YOLOv8 are both fine as long as the final mAP holds up under real-world conditions. Please apply best-practice augmentation, tune hyper-parameters, and track training with clear metrics so I can reproduce the results later in PyTorch. Deliverables • Curated and fully annotated vehicle image dataset (bounding-box labels in YOLO txt format) • Trained YOLO weights and configuration files • Short report summarising dataset composition, training settings, and evaluation scores ...
I’m...by SHAP values. I already have raw MRI images and basic labels; you would handle the full pipeline—from preprocessing and augmentation through model tuning, inference, and explanation generation. Code should be delivered in clean, modular Python (PyTorch preferred, but TensorFlow is fine if you make a strong case) with Jupyter notebooks for reproducibility. Acceptance criteria • Model must reach the benchmark accuracy we agree on during kickoff. • Grad-CAM overlays and SHAP plots render automatically for each study. • Outputs exportable as JSON for structured data and PNG for visuals. • README explains setup, training steps, and how to plug in new MRI data. If you have prior work combining SHAP and CAM techniques&m...
...library—as long as the final solution is reproducible on a single high-end GPU. Expect to work with a labelled dataset of roughly 100k images; if additional augmentation or cleaning pipelines are needed, include them in your approach. Accuracy matters, but I’m also watching inference speed, so mixed-precision training or model-quantisation tricks are welcome. Timeline is urgent: I’d like a first workable model within days, not weeks, and the polished hand-off shortly after. Deliverables • Python source code and environment file • Trained weight file(s) and clear instructions for re-training • A concise README explaining data prep, training commands, and inference calls • Brief performance report (accuracy, loss curves, infe...
I have a ~12 GB 1hr 15min video of a funeral service, that I would like to be edited with the following amends: - I want to trim the video at the start and end (I have timestamps that I want to use instead of the current length) - There are two points in the video where I would like to overlay two new video files. For context, in the 'main' funeral service video, you hear a song playing and can see a single shot view of a microphone. Instead, I would like to swap in a clip of that song which includes a photo montage that I have made (which was played on-screen at the funeral, but which you cannot see clearly in the footage). This is true for two songs. - I would like to apply some augmenting to the 'main' video: colouration, and touch-ups including (if this is possible...
...when a finger or phalanx is completely absent so that it can be recorded automatically I expect you will combine classical computer-vision preprocessing with a deep-learning model (PyTorch or TensorFlow are fine) and possibly leverage pose-estimation libraries such as MediaPipe or OpenCV for landmark detection. Accuracy matters more than perfection in lighting or background, so robust data-augmentation and domain-adaptation techniques will be key. Phase 1 Train and validate a model that ingests a single hand photograph and returns a JSON or CSV containing joint landmarks, the calculated angles, and a simple missing-phalanges Boolean for each finger segment. Phase 2 Package the model behind an easy interface where I can drop in one or many photos of the same patien...
We want to use the existing data and use other concepts for data augmentation compared to what we have now. Then we want to retrain the machine learning model.
...sharpen response accuracy so the system stays firmly grounded in fact. Right now the core pipeline is a Hugging Face Transformer model wrapped in a Retrieval-Augmented Generation (RAG) layer. I need you to audit the entire flow, diagnose where and why hallucinations appear, and then apply proven mitigation techniques. That could involve prompt engineering, better retrieval logic, truth-focused data augmentation, fine-tuning, or introducing guard-rail frameworks—whatever combination delivers measurably higher factual precision. Deliverables • A revised model or inference pipeline that demonstrably improves response accuracy (verified on my held-out benchmark). • Evaluation report with automatic metrics (e.g., factual consistency, BERTScore) plus a smal...
I need a small-scale AI model or script that reliably turns an input photo into a cartoon-style portrait with moderate detail. The focus is squarely on portraits, not landscapes or abstract work, and I’m after a playful yet recognisable look—clean...micro-web app with clear instructions • A brief README covering environment setup, dependencies and usage • Five sample outputs that demonstrate the expected moderate level of detail Acceptance criteria: given a 1024×1024 photo, the system produces a 1024×1024 cartoon portrait within 60 s on an RTX-class GPU, preserving likeness and meeting the stylistic brief above. Feel free to suggest data augmentation tricks or post-processing tweaks you find useful—I’m open to ideas that imp...
...EfficientNet, MobileNet, or a vision transformer—as long as the final model meets the accuracy targets we set together. Feel free to work in PyTorch or TensorFlow/Keras; I’m comfortable deploying either. What I’ll provide • A structured folder of training, validation, and test images • Category labels and a brief data dictionary • Access to a GPU instance if you need it What I need back 1. Clean, well-commented code (Jupyter notebook or Python scripts) that handles preprocessing, augmentation, training, and evaluation. 2. Trained weights plus an inference script that loads one or more images and returns the predicted class with confidence scores. 3. A concise report (Markdown or PDF) covering model architecture, key hyper-paramete...
...an end-to-end deep learning system that performs pixel-level flood segmentation on satellite imagery in real time. The model accurately identifies flooded areas from Sentinel-2 or similar multispectral satellite data and generates instant flood maps — ideal for disaster response, emergency management, agriculture monitoring, and urban planning. Key Highlights & Technical Achievements: Built a UNet-ResNet34 architecture that delivers high-precision binary and multi-class segmentation on satellite images. Designed a complete AI pipeline including data preprocessing, image augmentation, model training, real-time inference, and visualization of flood masks. Deployed the model as a production-ready REST API using FastAPI, enabling instant predictions via a sin...
...Saturn-inspired lamp shade that slips neatly onto the IKEA STRÅLA table lamp with its E14 socket. The shade will be printed in PLA, so wall thickness and overhangs must be optimized for single-material FDM printing without excessive support. Im aiming to improve the existing products , functionality, ergonomics and handleability by incorporating at least another 3D printing theme (specified below). The augmentation should be able to separate from the existing design. without damaging it. - Work in SolidWorks, or similar CAD, that’s perfect; just be sure the final file exports cleanly to STL (and ideally 3MF) and all assembly files Deliverables • Editable CAD file (native format) • Watertight STL/3MF ready for PLA printing • Brief print-orientat...
...suitable stack (Python, Scala, or another language you can justify) and libraries such as PyTorch, TensorFlow, scikit-learn, or Spark ML; just be prepared to explain why your choices fit a high-volume payments environment. Please base the model on real-world transactional datasets (synthetic augmentation is fine) and include a clear path for retraining as new data arrives. All code should be production-ready, containerised, and exposed through a REST or gRPC endpoint that my back-end can call. Deliverables • Data preprocessing scripts and reproducible training pipeline • Pattern-recognition model with inference service (Docker/K8s ready) • Unit and load tests proving stability at 1k TPS+ • Minimal latency benchmarks and optimisation not...
...current training loop intact. • Enhanced preprocessing – introduce smarter augmentation or normalization steps so the model sees cleaner, more diverse input. • Advanced post-processing – refine NMS or related filtering so final predictions score higher under my mAP criteria. I will share the code, the change log, and the benchmark sheet that shows today’s baseline. Your deliverable is the updated codebase plus a short report that compares before-and-after results using the same test set. Clean, modular commits and comments are important—this project will be handed over to other engineers later. If you’ve deployed similar accuracy lifts on image detectors before, especially by combining algorithm tweaks with data and post-proces...
...train/validation/test Augmentation must be applied only to the training split Validation and test must remain original, real, and unaugmented I do not want any synthetic or LLM-generated samples in validation or test The final evaluation setting must stay realistic and fair Augmentation requirement: To address class imbalance, you may use an LLM to generate augmented variants in a controlled way, but only under these conditions: augmentation must be applied after the split only the train split may be augmented augmented samples must not leak into validation or test validation and test should remain fully original and unchanged the process should improve training balance without making evaluation unrealistic What I need help with I want help designing and preparing...
...computer vision. Your role will involve understanding the original experiment, replicate and provide running code on the experiment, and implement modifications to improve results and fill the research limitation. If you have a passion for research and a proven track record in experimental methods, we would love to hear from you! Scope • Replication: mirror the author’s data preprocessing, hyper-parameters, and augmentation schedule until our accuracy matches their reported numbers (±1 %). • Explainability upgrade: introduce negative concepts explanation. • Packaging: clean, modular Python code plus a short README or notebook that walks through reproduction, visualisation, and how to plug in a new dataset. Acceptance criteria 1. Baseli...
I’m building a feature that takes a photo of a bedroom and instantly tells the user whether it contains a bed, wardrobe or nightstand. To keep the mobile app lean, I need a very lightweight computer-v...important, but compactness is equally critical, so let me know what trade-offs you recommend and past results you’ve achieved on similar lightweight object-detection tasks. When you reply, please outline: 1. Your preferred architecture and why it suits this job. 2. Expected final model size and typical inference speed on a mid-range phone or Raspberry Pi-class device. 3. Any data requirements or augmentation you’ll need from me. Once we agree on the approach, I’ll share a small curated dataset of bedroom images to get us started, and we can iter...
