PROJECT TITLE :

Feature-Based Lucas–Kanade and Active Appearance Models - 2015

ABSTRACT:

Lucas-Kanade and active look models are among the most commonly used ways for image alignment and facial fitting, respectively. They both utilize nonlinear gradient descent, that is usually applied on intensity values. In this paper, we tend to propose the use of highly descriptive, densely sampled image features for each problems. We have a tendency to show that the strategy of warping the multichannel dense feature image at each iteration is a lot of useful than extracting features once warping the intensity image at each iteration. Motivated by this observation, we have a tendency to demonstrate robust and accurate alignment and fitting performance employing a selection of powerful feature descriptors. Especially with the employment of histograms of oriented gradient and scale-invariant feature rework options, our method considerably outperforms the current state-of-the-art results on in-the-wild databases.