2023 updates

Update 2023-04-17

Added:

  • A pre-trained model for background removal from images
  • .webp format support

2022 updates

Major update 2022-03-22

Added:

  • Instance segmentation model training
  • AI-assisted labeling for instance segmentation
  • Pre-trained model for instance segmentation
  • Sample bitmap labels and sample segmentation model

Changed/improved:

  • Automatic removal of small bitmaps from the results of smart labeling tool

Update 2022-01-24

Added:

  • A ruler with XY coordinates in the labeling tool
  • Possibility to download Places Classification and OCR pre-trained models
  • Possibility to see labeled object counts for each project in project management screen

Changed/improved:

  • Updated OCR model
  • Improved loading speed for the sample project, and project management screen
  • Users are allowed to train multiple models at the same time

2021 updates

Update 2021-11-15

Added:

  • Possibility to see the average time it takes to label one image for each user
  • Possibility to duplicate a project
  • Various minor interface enhancements

Update 2021-10-04

Added:

  • Possibility to set roll, pitch and yaw angles in labeling tool
  • Thumbs up/Thumbs down evaluation of image labeling
  • Sample data set and sample models for new users
  • Possibility to label by image similarity via REST API
  • Possibility to change classification labels via REST API

Update 2021-08-23

Added:

  • Possibility to download General classification, Object detection and NSFW classification pre-trained models
  • Labeling by image similarity
    • Query photo is labeled based on N most similar photos where N can be set by user
    • Possibility to change score threshold for labels
    • Possibility to adjust labels manually
    • Possibility to save labeled images to your data set
  • Possibility to add classification labels from labeling tool window
  • Possibility to adjust brightness, contrast and saturation of the image
  • Possibility to save pose estimation results as keypoint labels
  • Possibility to train classification models without validation set
  • Right click context menus grouped into sub-menus

SentiSight.ai mobile app release 2021-07-21

View press release

SentiSight.ai mobile app allows to:

  • Make predictions using trained or pre-trained models
  • Use image similarity search function
  • Save predictions as image labels

SentiSight.ai mobile app can be downloaded on:

Update 2021-06-28

Added:

  • Project management window with the following functionalities:
    • Possibility to see a project list and a user list
    • Possibility to add or remove selected user(s) to/from selected project(s)
    • Possibility to add/rename projects and manage user roles and permissions
  • Possibility to see payment history for users
  • Improved performance

Major update 2021-05-17

Changed:

  • Subscription system will be replaced by new pay-as-you-go payment system. Read more about upcoming pricing here
  • Object detection model training will be available for free users
  • Changed classification and similarity search network
  • REST API requests using the “text/plain” content type are now considered as legacy. We recommend changing your implementations of our REST API to use the “application/json” content type instead. We also updated our code samples to take this change into account.

Added:

  • Possibility to retrain full classification model network (it is useful when you have a lot of images)
  • Possibility to stop classification model training if there is no improvement for a certain amount of time
  • Possibility to upload and delete images from a project using REST API
  • Added C# REST API code samples and Swagger specification

Update 2021-03-15

Added:

  • Pose estimation pre-trained model
  • “Redo” functionality in image labeling tool
  • A setting whether to disable smoothing when image is zoomed in

Changed:

  • Smart labeling tool
    • Users do not need to click “Extract” button, the extraction is done automatically
    • The active area can be changed or extended without losing progress

Update 2021-02-08

Added:

  • Text recognition pre-trained model
    • Possibility to parse text from images
    • 75 supported languages
  • Functionality to set a direction for bitmap masks
    • When setting a direction a bounding box will be fitted around a bitmap mask
    • The fitted bounding box can be rotated and an arrow will mark it’s direction
  • A setting to choose whether labeling outside image border is allowed

2020 updates

Update 2020-12-14

Added:

  • Possibility to open labeling tool in prediction screen

Changed:

  • Improved image annotation tool performance
  • Improved model train progress tracking

Update 2020-11-03

Added:

  • AI-assisted image labeling for classification and object detection
  • Possibility to review and correct AI-assisted labels
  • Possibility to filter images with AI-assisted labels

Update 2020-09-07

Changes:

  • Various performance improvements
  • Minor changes in the UI

Release of New Version 2020-08-17

View press release

Added free features:

  • Functionality to download offline models (30-day trial)
  • Possibility to use a pre-trained object detection model

Added paid features:

  • Paid licenses for offline models
  • Functionality to train object detection models
  • Possibility to share projects
  • Possibility to track labeling time of all users in a project
  • Functionality to filter images by people who have labeled or reviewed those images
  • Functionality to edit user roles and permissions
  • Possibility to have a dedicated image similarity search service

Update 2020-07-08

Added:

  • Possibility to upload color bitmaps (one bitmap per image)
  • Possibility to download/upload black and white bitmaps (one bitmap per object)
  • A color picker for object labels

Changed:

  • When a new object label is selected or created on canvas, the object list automatically scrolls to that object
  • Large file downloads are now prepared in the background
  • Added option to change brush size in the smart labeling tool
  • Brush preview is always displayed, even while drawing an object

Update 2020-06-09

Added:

  • Smart labeling tool for segmentation. Using this smart labeling tool users can quickly label images for segmentation by marking foreground and background in a selected area.
  • Functionality to convert bitmaps to polygons
  • Labeling time reporting. Users can now track how much time they have spent labeling day by day using our platform.
  • Download manager. Users can see their previous downloads and re-download them if necessary.
  • Possibility to label images in image similarity search result screen
  • Possibility to delete a specific label from all selected images at once
  • Possibility to delete all labels from all selected images at once
  • Possibility to hide all labeling objects

Changed

  • Now after labeling image for classification or exiting from labeling screen, the page is not reloaded, so users stay on the same page as they were before
  • Using pre-trained models does not require to specify project ID anymore

Update 2020-05-04

Added:

  • Three new pre-trained models that are ready-to-use without any additional training:
    • General classification – this model tells what is seen inside an image in general
    • Places recognition – this model tries to recognize the place where picture was taken
    • NSFW classification – this model can be used for content moderation purposes to filter out any images that contain nudity
  • Labeled objects now can have “attributes”. These attributes can be used to further describe the object. For example, object “table” can have attributes such as “white”, “round”, “four legs”, etc.
  • Images now have “seen” or “unseen” status. If a user opens an image for labeling, its status is automatically set as “seen”. In addition user can mark specific images as “seen” or “unseen” by simply selecting them and using right click context menu. Later the user can filter images that he has seen or not seen. This feature was implemented to help to track the labeling progress.

Changed:

  • Some minor changes image labeling interface (sorting labels newest-oldest, possibility to have empty label name field by default, etc.)

Update 2020-03-30

Added:

  • A completely redesigned labeling tool that includes bounding box, polygon, polyline, points and bitmap labeling using a brush tool
  • Functionality to mark keypoints for bounding box and polygon objects
  • Possibility to save keypoints template that includes names of the keypoints
  • Functionality to convert bounding boxes and polygons to bitmap (“rasterize”)
  • Possibility to see label names next to objects
  • Possibility to change the order of object list by dragging them
  • Possibility to select which type labels are downloaded
  • Possibility to filter images based on labels that are of other type than classification (such as polygons, bounding boxes, points and others)

Changed:

  • Now object detection and segmentation labeling are done within the same window
  • Some of the fields in the downloaded .json file with labels have changed. For some time we will support both the old and the new formats

Removed:

  • Fixed vertex count setting for polygon labeling
  • Continous polygon drawing mode
  • Full screen mode

Update 2020-01-27

Added:

  • Object name is shown next to the bounding box or polygon
  • Image labels are now downloaded as a .zip file and works faster
  • Now we report both label-wise and image-wise accuracy in multi-label classification
  • Possibility to upload classification labels from .json file (rather than .csv)
  • Possibility to rotate image counter-clockwise

2019 updates

Update 2019-11-20

Added:

  • Possibility to download image similarity search results
  • Improved object detection/segmentation labeling interface:
    • Bounding boxes and polygons that are being edited are highlighted in an image and in the list on the right panel
    • Active vertices are highlighted
    • A new vertex which is about to be added on an edge is visualized
    • After finishing labeling a bouding box or a polygon its name field is automatically activated
    • Occluded objects are marked by a dashed line
    • New shortcuts for labeling (‘O’ – mark object as occluded, ‘V’ turn on/off visibility of an object)
  • Added a search field by image filename
  • A new button to select all images in the current page
  • Five new video tutorials (training a model on one class, image similarity search, labeling images for object detection, labeling images for segmentation, labeling settings)

Changed:

  • Removed “Move” and “Edit” modes in image labeling (everything is done in a single mode now)
  • Restructured image similarity search results window
  • Slightly changed REST api output format
  • User projects with duplicated names were renamed
  • Download labels function now downloads classification labels in .csv file in addition to all types of labels in .json file
  • When a user exits labeling, the current page stays where the last image was and the last image is highlighted

Update 2019-09-20

Added:

  • New shortcuts for labeling (‘Z’ to fit image to screen, ‘Ctrl or Shift + mouse wheel’ to scroll image left/right or up/down)

Changed:

  • A minimum value of optimized score threshold for multi-label classification is now 5%
  • Downloaded image labels for object detection, segmentation and classification are now all in one .json file
  • Reordered context menu of right mouse button

Update 2019-08-30

Added:

  • Functionality to perform image similarity search via web interface and REST api
  • Possibility to see how many images will be allocated to train and validation sets
  • Functionality to rotate image
  • Functionality to sort images by name or upload time
  • Possibility to select images from several different pages
  • New shortcuts: press ctrl+1, ctrl+2, etc. to select object label
  • New shortcut: delete images with [delete] key
  • Possibility to pre-process images for classification and similarity search during image upload

Changed:

  • Restructured project selection and editing menu
  • Removed boundary groups from object detection labeling (now each bounding box is stand-alone)

Update 2019-06-26

Added:

  • Functionality to upload object detection and segmentation labels from a .json file
  • Functionality to select and move polygons by mouse click during labeling
  • Full screen mode for image labeling
  • Icons to mark images labeled for object detection and segmentation
  • Possibility to exclude labels from training in the training screen
  • Possibility for changing line width for labeling
  • Functionality to filter images by labels using “and” and “or” filters
  • Functionality to upload images by selecting a whole folder

Changed:

  • Removed “save” button from settings menu (now settings are saved automatically)
  • Now images in “show predictions” screen are shown in original aspect ratio
  • Restructured upper menu

2018 updates

Updates 2018-11-18 – 2019-06-26

Added:

  • Possibility to upload classification labels as a .csv file
  • Possibility to download classification predictions as a .csv file, .json file or images grouped into folders by predicted labeling
  • Download object detection and segmentation labels as a .json file
  • Upload images as a .zip file with a possibility to label them based on folder structure
  • Possibility to select which exact images will be used for validation
  • Possibility to set “default” label on all selected images
  • Possibility to have a fixed bounding box ratio
  • Possibility to have a fixed number of points in a polygon
  • Possibility to skip images that are already labeled
  • Possibility to switch between images using arrows or “A” and “D” shortcuts

Changed:

  • Labels without the minimum required number of images are automatically excluded from training
  • Improved bounding box and segmentation labeling tools

SentiSight.ai first release 2018-11-18

View press release

Added:

  • Functionality to train single-label and multi-label classification models
  • Functionality to label images for classification, object detection and segmentation
  • Functionality to view predictions on train and validation sets and analyze model’s performance
  • Functionality to filter images by type and labels
  • REST api service