下载中心

MVTEC Deep Learning Tool 24.12 Full Version x64 windows 深度学习工具dlt-24.12完整版

文件名: dlt-24.12.zip
文件大小: 4268138142 字节 (3.98 GB)
修改日期: 2024-12-18 22:40
MD5: e4bed21db2a896c4036dbe20bb9445be
SHA1: ff1614516f21dd225b0680586146f804ab6d4278
SHA256: 1c9c624c23868e8c1ab8300a9338056c57b0e9045c8be9b21b0f28d3588277d5
CRC32: 095c1624

MVTEC官方地址：
需要注册，然后免费下载。www.mvtec.com

百度网盘下载地址：

MVTEC Deep Learning Tool 24.12 Full Version x64 windows 深度学习工具dlt-24.12完整版

http://visionbbs.com/thread-34045-1-1.html

(出处: LabVIEW视觉)

New features

• DLT now supports training and evaluation of Deep OCR models. Detection and recognition can be trained separately.

• The control panes on the right part of the Image page are now collapsible. The behavior of all collapsible panes was unified so that it is now possible to unfold or collapse the panes also via the splitter handle or by clicking on the title.

• The underlying model used by the Smart Label Tool – Hover & Click has been updated. Now, the necessary preprocessing per image is faster.

• To improve training in very rare cases for Anomaly Detection projects, the option to set the advanced training parameter 'Weight Prior' has been added.

• The buttons to start, resume, or stop a training or an evaluation are now part of the respective entry in the training or evaluation list.

• The DLT is now based on HALCON 24.05.

• The icon for 'Estimate advanced network parameters' on the Training page has been changed.

• It is now possible to switch between labels by pressing Tab (forward) or Shift+Tab (backward). This also applies for switching between text boxes in Deep OCR projects.

• When switching between different evaluations on the Evaluation page, their settings are now persistent as long as the project is open. This allows better comparison of evaluations.

• For Deep OCR projects, the character distribution of labeled images can now be shown in the statistics dialog.

• It could happen that the DLT could not be started if another application was installed that used an incompatible version of OpenVINO. This problem has been fixed. Now, the OpenVINO version used by the DLT no longer interferes with other OpenVINO installations.

• For Classification projects, it is now possible to visualize an alternative, Guided Grad-CAM-based heatmap.

  
  

Resolved issues and improvements

• When resizing the Global Measures area on the Evaluation page, the shown pie charts were scaled wrongly and could become much too small. This problem has been fixed.

• The alignment of the zoom-in and zoom-out cursors of the Class Distribution List on the Split page sometimes was not optimal. This problem has been fixed.

• In the evaluation reports for instance-based projects, the column for total number of predictions and the row for total number of GT labels were missing. This problem has been fixed.

• The navigation between pages with Alt+<Page Number> was broken. This problem has been fixed.

• Since DLT 24.05 using HALCON 23.11, the required CUDA version is 12.1. The corresponding information in the documentation was out of date. This problem has been fixed.

• Zooming into images via the mouse wheel sometimes also unexpectedly scrolled the image. This problem has been fixed.

• When evaluating a training of an object detection project for oriented rectangles via the Evaluation page, the postprocessing parameters that can be set on the Settings > Advanced Inference Parameters pane were not used. This problem has been fixed.

• When activating the Smart Label Tool – Hover & Click, it could happen that the DLT crashed when going through the images. Further, sometimes the preprocessing bar did not show a consistent state. These problems have been fixed.

• In Deep OCR projects, the suggested text was sometimes not cleared when pressing a key. This problem has been fixed.

• In some cases, the export setting of the HALCON 20.11 compatibility mode was ignored. This problem has been fixed. Further, now the compatibility mode is not available if optimization for an AI² device is selected. In this case, the checkbox is not shown, and the mode cannot be chosen.

• The Hailo WSL installation script now supports spaces in paths.

• In the Chinese translation of the documentation, an entry in the “Minimum System Requirements” table was missing. This problem has been fixed.

• For Deep OCR projects, it was possible to paste a string containing non-printable characters (including , , etc.) into the field for the recognized text. This problem has been fixed. Now, all non-printable characters are removed from the pasted string. Additionally, an option has been added to the global preferences dialog allowing this behavior to be changed.

• The status of the Heatmap button on the Evaluation page could be misleading if a project was opened while the list of available DL devices was not ready. This problem has been fixed.

• The DLT could crash when using a touch device and interacting with label regions. This problem has been fixed.

• After optimizing a model for TensorRT or OpenVINO, then switching to Hailo and changing the setting for the option 'Use allocation script', the prior optimizations seemed to be lost. This problem has been fixed.

• In GC-AD projects, when drawing a polygon to restrict the domain for postprocessing, the lines were invisible during the drawing. This problem has been fixed.

• When optimizing a model for Hailo, there is now an option that allows enabling a fine-tuning step during optimization. This can improve the performance of the optimized model with the trade-off of a longer optimization time.

• In some cases, the placeholder text in the comment area was not located at the right position. This problem has been fixed.

• If in a GC-AD project from a combined trained model only the local or global subnetwork was selected for optimization, this selection was ignored, and always the combined network was optimized. This problem has been fixed.

• When duplicating a training during a running optimization, the wrong optimization state was shown in the duplicated training. This problem has been fixed.

• If an error occurred during the inference calculation, e.g., due to a missing image, it could happen that the error message was not announced properly. In some rare cases, the DLT could even crash. These problems have been fixed.

• Several clicks on the Duplicate entry in the training item context menu could create several duplicates of the training. This problem has been fixed.

• When selecting images or items on the Evaluation page, unintuitive effects could occur. This problem has been fixed.

• In Semantic Segmentation projects, the optimization of trained models for OpenVINO with precision fp16 failed in most cases. Hence, this option was removed from the precision selection box. Further, in some cases, the precision selection box contained unsupported or not all supported values. These problems have been fixed.

• If the learning rate was changed in the text box and afterward, another training parameter (e.g., number of iterations) was changed using the spin box or mouse wheel, the learning rate changed back to its old value. This problem has been fixed.

• The documentation now provides a more detailed description of anomaly score and anomaly score tolerance.

•