UCLA researchers propose PhyCV: a physics-inspired computer-vision Python library

Artificial intelligence is making noteworthy strides in the field of computer vision. One major area of ​​development is deep learning, in which neural networks are trained on huge data sets of images to recognize and classify objects, scenes, and events. This has resulted in significant improvements in image recognition and object detection. Combining computer vision with other technologies opens various gateways to new possibilities and scopes for artificial intelligence.

In the latest innovation, Jalali-Lab @ UCLA has developed a new Python library called PhyCV, which is the first physics-based computer vision Python library. This unique library uses algorithms based on the laws and equations of physics to analyze the imaging data. These algorithms simulate how light passes through many physical materials and are based on mathematical equations rather than a series of handcrafted rules. The algorithms in PhyCV are built on the principles of a rapid data acquisition method called Photonic Time Stretching.

The three algorithms included in PhyCV are – Phase Stretch Transform (PST) Algorithm, Phase Adaptive Gradient Extractor (PAGE) Algorithm, Visibility Enhancement via Virtual Diffraction and Coherent Detection (VEViD) Algorithm.

Phase span transform (PST) algorithm

The PhyCV library’s PST algorithm identifies edges and textures in images. The algorithm simulates how light travels through a device with certain diffraction characteristics and then coherently detects the subsequent image. The algorithm works best with images with visual impairments and has been used in various applications, including improving the accuracy of MRI scans, identifying blood vessels in retinal images, etc.

Adaptive Gradient Field Extractor (PAGE) Algorithm

The PAGE algorithm determines edges and orientations in images using principles of physics. Essentially, PAGE simulates the process of light passing through a device with a defined refractive structure, which results in the conversion of an image into a complex function. Information about edges is stored in the real and imaginary components of the result. The researchers mention how PAGE can be used as a pre-processing method in various machine learning problems.

Visibility improvement via the VEViD algorithm

The VEViD algorithm improves low-light and color images by considering a spatially variable light field and using physical processes such as diffraction and coherent detection. It does this with minimal latency and can therefore increase the accuracy of the computer vision model in low-light conditions. A specific approximation of VEViD, known as VEViD-lite, can enhance 4K video at up to 200 frames per second. The research team compared the VEViD algorithm with popular neural network models showing how VEViD shows exceptional image quality with processing speed greater than just one to two orders of magnitude.

PhyCV is available on GitHub and can be easily installed via pip. Algorithms in PhyCV can even be implemented in actual physical devices for more efficient computation. PhyCV undoubtedly sounds interesting and looks like an important development in the field of computer vision. Thus, advances in artificial intelligence and computer vision are certainly driving a wide range of advanced applications.

scan the github And The project. All credit for this research goes to the researchers on this project. Also, don’t forget to join Our Reddit pageAnd discord channelAnd Email newsletterwhere we share the latest AI research news, cool AI projects, and more.

Tania Malhotra is a final year from University of Petroleum and Energy Studies, Dehradun, and is pursuing a BTech in Computer Science Engineering with a specialization in Artificial Intelligence and Machine Learning.
She is passionate about data science and has good analytical and critical thinking, along with a keen interest in acquiring new skills, leading groups, and managing work in an organized manner.

Leave a Comment