Q: What are some simple ways people can get control of their data?
A: Data for computer vision is available in every industry where there is the use of cameras or any other form of imaging equipment – be it optical images, thermal images, X-rays, or even satellite imagery. The training and evaluation of deep learning models applied for computer vision requires us to provide good quality datasets with various labels. To get control of this data, the company data management personnel need to develop a standardized taxonomy and data labeling schema. The images or videos need to be co-located as raw bytes with the metadata of the image. The metadata can then provide a taxonomy for an image classification problem and more complex and time-consuming rectangular or polygonal annotations for object detection, recognition, and segmentation problems.
Q: Can you give an example of a use case of how using computer vision helps companies make better use of their available data?
A: One example would be cameras that are mounted in most offices, stores, or workshops with the video feed used as a surveillance application. It can help companies keep track of unauthorized personnel entering the monitored facility, spot fishy behavior or unusual activities in the building to ensure safety and prevent theft, averting shoplifting, etc. Surveillance components are pictorial so diagnosing and cross-examination of these components is time-consuming, plus the maintenance is expensive. Computer vision helps overcome these issues.
Q: What’s the coolest thing you have seen with a computer vision application?
A: With advancements in GPU computing technology available today, Generative adversarial networks popularly known as GAN’s have made it possible to create beautiful paintings. Advanced applications of GAN’s can be used in video prediction, specifically predicting up to a second ahead of video frames with success, mainly for static elements of the scene.
Q: What industries do you see embracing computer vision, and which industries are not but should?
A: The autonomous car industry is one of the major reasons for the advancement of computer vision technology. The other industry advancing in computer vision is healthcare, where most data is image-based. Several other industries behind the two listed above embracing computer vision applications are sports, agriculture and manufacturing all to make better, more intelligent decisions.
In my opinion, smaller retail chains have an enormous opportunity for using computer vision applications. Surveillance is only the beginning, Amazon Go is an excellent example where customers can shop at a store, the camera tracks what the customer selects, and then the cost of the item is automatically charged to the customer’s account.