Abstract: Handcrafted audio descriptors and learned deep representations each bring distinct strengths and inherent limitations to speech emotion recognition (SER). Traditional handcrafted features ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results