SentiSight: Machine Learning Platform for Image Recognition
SentiSight supports businesses in retail, manufacturing, and e-commerce in developing custom image recognition models. It is designed for teams that need to label visual data and deploy models via REST API or offline.
At a glance
- Category
- Software Development
- Best for
- Retail and e-commerce businesses, Manufacturing operations, Research and education institutions, Software development teams
- Pricing
- SentiSight uses a pay-as-you-go wallet system. Costs include €1 per 1,000 predictions, €1 per 1,000 labels, and €3.6 per hour of training, with regressive pricing for higher volumes. New users receive €20 in free credits and €5 monthly complimentary credits.
- Key use cases
- Retail Product Tagging, Manufacturing Quality Control, E-commerce Visual Search, Content Moderation, Text Extraction
- Official website
- sentisight.ai
SentiSight.ai is a web-based platform designed for image labeling and the development of AI-based image recognition applications. It allows users to annotate images and train deep neural network models within a single environment, supporting an iterative workflow where labels can be refined and models retrained.
The tool is designed for organizations in research, education, retail, and manufacturing. It provides tools for image classification, object detection, and segmentation, catering to users who prefer a no-code interface as well as experts who may customize training parameters.
Buyers can use pre-trained models for immediate tasks or build custom models. The platform supports various deployment methods, including a web interface, REST API, and downloadable offline models for local use.
Buyers should confirm their technical environment for offline deployment, as offline models require a Linux system and an NVIDIA GPU for maximum speed.
Key Features
Supports bounding boxes, polygons, points, polylines, and bitmaps for annotating visual data.
Supports training for image classification, object detection, and segmentation models.
Provides ready-to-use models for background removal, NSFW detection, and OCR in 75 languages.
Supports 1vN and NvN searches to find visually similar images within a dataset.
Designed to help extract objects by marking foreground and background points to assist with bitmap labeling.
Allows the integration of trained models into external applications for predictions.
Use Cases
Automating the tagging of products to support search results and reduce manual entry.
Using object detection to help identify product inconsistencies and imperfections.
Implementing image similarity search to help customers find visually similar items.
Using pre-trained NSFW models to help identify inappropriate content.
Using a pre-trained OCR model to recognize and extract text in 75 different languages.
Best For
Pricing
SentiSight uses a pay-as-you-go wallet system. Costs include €1 per 1,000 predictions, €1 per 1,000 labels, and €3.6 per hour of training, with regressive pricing for higher volumes. New users receive €20 in free credits and €5 monthly complimentary credits.
FAQ
It uses a pay-as-you-go wallet system with costs for predictions (€1/1000), labeling (€1/1000), and training (€3.6/hour). New users receive €20 in free credits, and all users get €5 monthly.
Yes, models can be downloaded for offline use. A 30-day free trial is available for offline models, after which a license must be purchased.
The platform is designed to allow users to implement deep learning techniques interactively without coding, though REST API access is available for developers.
Source category: Software Development
Source subcategory: Computer Vision
