Author: Philip Godfrey
What is OCI Vision?
OCI Vision serves as an advanced AI tool aimed at facilitating extensive image analysis powered by deep learning algorithms. It streamlines the process of integrating image and text recognition features into applications, eliminating the need for specialised knowledge in machine learning.
What does OCI Vision do?
- Provides pretrained and customisable computer vision models to analyse image-based content.
- You can upload images to detect and classify objects in them.
- Complete integration with OCI Data Labelling simplifies data labelling.
- OCI provides access to pretrained models, enabling users to develop custom models by uploading and training image classification models.
- Subsequently, users can assess the metrics of custom models and deploy them seamlessly within Oracle Analytics Cloud for analysis of new image datasets.
This facilitates rapid innovation and seamless integration of results into an analytics platform. The combined utilisation of OAC and OCI Vision offers a versatile solution applicable across various industry use cases.
Pretrained Models:
- Image Classification
- Object Detection
- Text Recognition
Custom Models:
- Image Classification
- Object Detection
|
Feature
|
Description
|
Common Applications / Examples
|
|---|---|---|
|
Optical Character Recognition (OCR) & Text Extraction
|
Extract textual features from documents, such as printed and handwritten documents
|
• Business reports in PDF form • Receipts and invoices • Insurance or other application forms
|
|
Table Extraction
|
Extract text in tabular format with row/column information per cell.
|
• Item-level details in receipts • Invoices • Purchase orders
|
|
Key-Value Extraction
|
Extracts key-value information based on a pre-trained model of common document types
|
• Invoices • Receipts • Passports / Driver Licence
|
|
Document Classification
|
Classifies documents into different types based on high-level features and extracted keywords.
|
• Receipt • Resume •Driver Licence
|
What are some use cases for Vision?¹
Insightful Analysis from Visual Data:
Leverage AI Vision to automatically extract textual or visual information from images, empowering analytical applications. By amalgamating visual data with supplementary inputs like product sales history and customer reviews, gain comprehensive insights into trends such as product availability and popularity.
Efficient Digital Asset Management:²
Enhance image-based files within Oracle Cloud Infrastructure AI Vision by enriching them with metadata, including object identification and colour classification. This facilitates superior indexing and retrieval within digital asset management systems or expansive data warehouses, enabling seamless image retrieval for diverse applications ranging from intelligent search to retail management.
Anomaly Detection in Visual Data:³
- Utilise AI Vision to classify products or equipment as standard or defective based on visual cues like discoloration, tears, rust, deformities, or breaks. Automate the detection of anomalies to promptly flag defective materials, thereby expediting the need for repairs and maintenance.
Scene Monitoring and Object Localisation:
Leverage AI Vision to monitor scenes and identify specific elements within images. Detect the growth of vegetation near power lines or ascertain the availability of trucks at delivery lots by accurately locating objects and entire scenes within images. Whether localising particular objects within a scene or classifying the entire image, AI Vision offers precise scene analysis capabilities.
How to use AI Vision?
Interested to know more?
We’d be more than happy to discuss your requirements or questions, please get in touch here.
[1] https://www.oracle.com/uk/artificial-intelligence/vision/
[2] OCI Vision Overview – Exploring the service (Oracle Developers): https://www.youtube.com/watch?v=eyJm7OlaRBk
[3] Anomaly detection in images (Oracle Learning): https://www.youtube.com/watch?v=xy1T_3BeO4o
[4] https://www.oracle.com/uk/artificial-intelligence/vision/
Contact us today to arrange an assessment or email:
Philip Godfrey
Principal Data Analytics & AI Consultant at Vertice
