In this computer vision training you learn to prepare image data for models, tell classification, object detection and segmentation apart and measure model quality robustly. It is aimed at experienced developers and ML engineers. You work with torchvision, augmentation and metrics such as IoU and mAP, and analyse failure cases systematically. Python and deep learning knowledge are required. We adapt data examples and scope to your use case when you enquire.
What you’ll take away.
- Structure and annotate image datasets and split them into training, validation and test parts
- Apply image preprocessing and augmentation with torchvision transforms deliberately
- Distinguish classification, object detection and segmentation by task and output format
- Evaluate models with confusion matrix, IoU and mean average precision
- Sort failure cases by cause, such as data gaps, label errors or lighting, and derive countermeasures
Your training in detail
- Who is it for?
- Developers, ML engineers and data scientists who bring image analysis into products or processes, such as quality inspection or document processing.
- Prerequisites
- Python, basic deep learning knowledge (training, loss, overfitting) and experience with PyTorch or a comparable framework.
- Format & planning
- Suggested format: 3 days live online or on site, split into modules. Scope depends on prior knowledge.
Agenda: from understanding to application
Our independently developed agenda proposal. Focus areas, lab environment and scope are agreed before booking.
Understanding image data: pixels, colour channels, resolution, formats and typical quality problems.
Data preparation with annotation, label quality, class imbalance and leakage-free data splits.
Augmentation and preprocessing with torchvision transforms such as cropping, flipping, colour jitter and normalisation.
Models for classification, object detection and segmentation, and transfer learning with pretrained networks.
Evaluation with confusion matrix, precision and recall, IoU, mAP and thresholds.
Error analysis: forming error classes, reviewing examples, closing data gaps and testing robustness to lighting or camera changes.
Suggested exercise: You fine-tune a pretrained network on a small image dataset, compute the confusion matrix and sort the 30 most frequent misclassifications by cause in a table.
Exams, licences & credentials
There is no single authoritative external exam for computer vision. Attending alone awards no vendor certificate. Exam fees and licences are only included if the offer confirms them.
Vendor references & review date
Sources checked on 2026-11-03. These sources describe the technology or external exam requirements. They do not imply a partnership with NextNowa.
Your learning content
These objectives describe the training topic. The specific syllabus, duration, dates, trainer and price are agreed before booking.
Frequently asked questions
How many images do I need for a computer vision model?
There is no fixed number. With transfer learning, a few hundred well-labelled images per class are often enough for simple classification, while demanding detection usually needs more. Quality, variety and label cleanliness are decisive. You learn how to estimate the need from learning curves.
What do IoU and mAP mean in object detection?
IoU measures how much the predicted box overlaps the true one: intersection area divided by union area. The mAP summarises precision across recall levels and classes, often at several IoU thresholds. Together they show whether a detector finds objects and localises them precisely.
Can I use a pretrained model directly?
Often yes, as a starting point. Pretrained networks provide good features but usually need adapting to your images, such as camera, lighting or object classes. Also check the model's licence and training data. In the training you fine-tune such a model and measure whether it suits your case.
Does Computer vision include a certificate?
There is no single authoritative external exam for computer vision. Attending alone awards no vendor certificate. Exam fees and licences are only included if the offer confirms them.