Zum Inhalt springen

Computer Vision: Object Detection and Segmentation with Mask R-CNN

KI und Computer Vision

Computer Vision: How to Do Object Detection and Segmentation with the latest Mask R-CNN Algorithm

Mask R-CNN is a Deep Learning method for computer vision systems. It extends Faster R-CNN and adds a branch for predicting an object mask in parallel with the existing branch for bounding box recognition. Mask R-CNN is simple to train and adds only a small overhead to Faster R-CNN, running at 5 fps. Moreover, Mask R-CNN is easy to generalize to other tasks. For exmaple it allowing us to estimate human poses in the same framework. Mask R-CNN outperforms all existing, single-model entries on every task, including the COCO image dataset. The simple and effective approach is described in this research paper. Code example can be downloaded on Github DETECTRON. It can be used as a solid baseline and help ease future development in instance-level recognition.

The COCO dataset contains 80 object categories:

Computer Vision Imaging Dataset
Source: cocodataset.org

[bctt tweet=”Computer Vision: Object Detection and Segmentation with Mask R-CNN #ComputerVision #AI #ArtificialIntelligence #ML #DeepLearning #TransferLearning #ObjectSegmantation #RCNN #Industrie40 #Industry40 ” username=”AISOMA_AG”]

The underlying demo video was shot in Frankfurt am Main (Germany). Hope you like it.

 

Current article on the subject of computer vision and industrial image processing:

Computer Vision Use Cases
AI für Smart Cities