Deep learning basics and practical approaches with IDS NXT ocean
This video introduces the fundamentals of deep learning and demonstrates how to approach real‑world machine vision tasks using IDS NXT ocean. It explains key concepts such as neural networks, feature extraction and convolutional models, and shows how these methods can be applied effectively without requiring programming expertise.
You’ll learn how to prepare high‑quality datasets, avoid common pitfalls such as bias and imbalance, and iteratively improve your model using the IDS NXT lighthouse training platform. The video also walks through transfer learning, model selection, augmentation and deploying trained networks directly on IDS NXT cameras. Whether you’re new to deep learning or looking to refine your workflow, this session provides a clear, practical guide.
Topics covered
- Deep learning terminology and core concepts
- How to build high‑quality datasets for machine vision
- Transfer learning and training workflows in IDS NXT lighthouse
- Model selection, augmentation and avoiding overfitting
- Deploying neural networks on IDS NXT cameras
- Practical examples using presence/absence and error detection
Video timeline
- 00:00 - Introduction
- 01:30 - IDS NXT ocean introduction
- 02:48 - Agenda of this presentation
- 04:16 - Disambiguation
- 06:08 - Learning the rules (basic part)
- 15:09 - Approach deep learning problems (practical part)
- 17:45 - IDS NXT ocean workflow
- 26:05 - Image quality
- 28:02 - Data quality
- 34:32 - The training pipeline
- 40:56 - Transfer learning and fine-tuning
- 43:06 - Model accuracy and speed
- 49:32 - Summary
Who should watch
Ideal for machine vision engineers, automation specialists, AI newcomers, and developers who want to implement deep learning efficiently using IDS NXT ocean.
Watch now to master the essentials of deep learning with IDS NXT ocean.
