Details

Autor(en) / Beteiligte

• Cavazza, Jacopo
• Zunino, Andrea
• San Biagio, Marco
• Murino, Vittorio

Titel

Kernelized covariance for action recognition

Ist Teil von

• 2016 23rd International Conference on Pattern Recognition (ICPR), 2016, p.408-413

Ort / Verlag

IEEE

Erscheinungsjahr

2016

Quelle

IEEE Xplore

Beschreibungen/Notizen

• In this paper we aim at increasing the descriptive power of the covariance matrix, limited in capturing linear mutual dependencies between variables only. We present a rigorous and principled mathematical pipeline to recover the kernel trick for computing the covariance matrix, enhancing it to model more complex, non-linear relationships conveyed by the raw data. To this end, we propose Kernelized-COV, which generalizes the original covariance representation without compromising the efficiency of the computation. In the experiments, we validate the proposed framework against many previous approaches in the literature, scoring on par or superior with respect to the state of the art on benchmark datasets for 3D action recognition.