Robovision AI 5.3 release notes¶

• Robovision AI 5.3.1
• Robovision AI 5.3

Robovision AI 5.3.1¶

Release date: October 31, 2023

What's new¶

To streamline the annotation process and offer more control over members' collaborative efforts, we have enhanced the annotation curation. This includes:

• Copying and replacing annotations

  You can now copy or replace your own annotations with annotations from another source—another member's annotations or annotations generated by the Prediction tool. For more information about the functionality, see Copy or replace annotations.

• Deleting annotations

  You can now effortlessly delete your own annotations, annotations made by other members, and predicted annotations. For more information, see Delete labels.

Bug fixes¶

Label center

• Multi-Label EfficientNet: You can now batch delete labels from multiple samples in the Selected panel, in the thumbnails and list views of the label center.
• We have resolved an issue where 3-digit numbers in the confusion matrix chart were unreadable due to compression. The update ensures proper spacing, enhancing the overall readability of numerical data in the chart.

Test center

• We have optimized the loading of the test center to ensure smoother and quicker rendering of the page.

Known limitations¶

General

• The Robovision AI UI has been optimized for the browser size of 1920x1200.
• You will be logged out of the platform if your data import has been in the Uploading stage for more than an hour.
• If the storage is full, some functions may become unavailable, for example, saving annotation or training a model.

Browser support

• The Robovision AI platform has been designed and validated for Google Chrome 85 and Mozilla Firefox 77 or later.
• In Google Chrome, training metrics and a final validation score might not be displayed if an ad blocker is used. To prevent this, mark the Robovision AI URL as trusted in the Google Chrome settings.
• In Google Chrome, the thumbnail view can load slowly if you have password managers like Lastpass activated. To prevent this, disable the extension or work in a guest profile window.

Backward compatibility

• Backward compatibility with Robovision AI versions 3.x and 4.x is not foreseen. But upon request, Robovision will provide the needed assistance and conversion scripts.

Images

• Types supported: In general and with respect to the model testing functionality, the following image types are supported: JPEG, JPG, PNG, TIFF, BMP, GIF, including a 4^th transparency channel. EXIF orientation metadata is supported.
• File names must contain only Latin (ASCII) characters. There are known rendering, packing, and backup/restore issues with file names/object keys that contain non-Latin characters.

Label center

• The test predictions are displayed in the label center as the labels by Unknown.
• 2D labeler performance (in particular the Magnetic lasso and Grab cut tools) may degrade for high-resolution images.
• You may face the following issues when using the Prediction tool in the label center:
  • The prediction process gets hidden if you leave the label center. To view it again, click the Prediction tool.
  • The predicted labels are not added to the samples as the prediction process is running.
  • The Status section is not updated as the prediction process is running.

Train center

• During training setup, the Add model button to select a model for transfer learning is enabled even if there is no model available.
• Empty annotations are not displayed in the class distribution for training data even if training data contains empty annotations.

Inference

• Multi-Label EfficientNet: The Confidence threshold parameter is not displayed during the inference setup.

Robovision AI 5.3¶

Release date: October 3, 2023

What's new¶

Resetting inference parameters

• You can now reset inference parameters to the default ones when the inference has been stopped.

Improved tags functionality

• To make tags easier to use, we have moved them below the classes in the left panel. That’s where you can create, edit, and assign classes. To filter by the assigned classes, check the Filters panel on the right. For more information about tags, see Group data with tags.

Job management

• You can now view all background jobs started in different projects on a single page. For more information about job management, see Job management.

About page

• We have replaced the Settings page with the About page. On the About page, you can do the following:

  • Check the installed Robovision AI version and installed algorithms.
  • Change the platform color theme.
  • View the license information.
  • Download server logs.
  • View the support information.
  • View and manage running processes.

Bug fixes¶

Label center

• To prevent the creation of datasets with 0 samples, we now inform the user about the invalid train/validation split.
• When deleting labels in the single view of the label center, you will now have to confirm the deletion.
• The issue with the Grab cut tool getting stuck in an invalid state has been fixed.

• When opened in full-screen, the class distribution chart now shows:

  • Samples that are not annotated.
  • Samples with empty annotations.
  • Class names.
  • Legend indicating the count of samples or annotations.

• All algorithms now support MONO8 images.

• You can now use Shift in the list view to select/deselect samples.

Licensing

• We now support licenses with the infinite duration (for example, when no expiration date was set).
• The Edge license information in now aligned to the license information of the connected Robovision AI instance.

Train center

• We have fixed the issue where training would frequently freeze when training with large datasets.
• On the Google Cloud deployments, jobs like training, testing, or predictive labeling wouldn't start if the GPU toggle was switched off. This issue has been fixed.

Test center

• Test validation results now include samples that do not have labels.
• You can now give custom names to your test runs.

Others

• The number of available GPUs on the About and All resources pages is not valid in autoscaling scenarios which include the Google Cloud deployments.

Known limitations¶

General

• The Robovision AI UI has been optimized for the browser size of 1920x1080.
• You will be logged out of the platform if your data import has been in the Uploading stage for more than an hour.
• If the storage is full, some functions may become unavailable, for example, saving annotation or training a model.

Browser support

• The Robovision AI platform has been designed and validated for Google Chrome 85 and Mozilla Firefox 77 or later.
• In Google Chrome, training metrics and a final validation score might not be displayed if an ad blocker is used. To prevent this, mark the Robovision AI URL as trusted in the Google Chrome settings.
• In Google Chrome, the thumbnail view can load slowly if you have password managers like Lastpass activated. To prevent this, disable the extension or work in a guest profile window.

Backward compatibility

• Backward compatibility with Robovision AI versions 3.x and 4.x is not foreseen. But upon request, Robovision will provide the needed assistance and conversion scripts.

Images

• Types supported: In general and with respect to the model testing functionality, the following image types are supported: JPEG, JPG, PNG, TIFF, BMP, GIF, including a 4^th transparency channel. EXIF orientation metadata is supported.
• File names must contain only Latin (ASCII) characters. There are known rendering, packing, and backup/restore issues with file names/object keys that contain non-Latin characters.

Label center

• 2D labeler performance (in particular the Magnetic lasso and Grab cut tools) may degrade for high-resolution images.
• You may face the following issues when using the Prediction tool in the label center:
  • The prediction process gets hidden if you leave the label center. To view it again, click the Prediction tool.
  • The predicted labels are not added to the samples as the prediction process is running.
  • The Status section is not updated as the prediction process is running.

Train center

• During training setup, the Add model button to select a model for transfer learning is enabled even if there is no model available.
• Empty annotations are not displayed in the class distribution for training data even if training data contains empty annotations.

Inference

• Multi-Label EfficientNet: The Confidence threshold parameter is not displayed during the inference setup.