Using sparse representations for exemplar based continuous digit recognition
Jort Florent Gemmeke, Louis ten Bosch, Lou Boves, Bert Cranen · Lirias · 2009
A BSTRACT This paper introduces a novel approach to exemplar-based con nected digit recognition.The approach is tested for different sizes of the exemplar collection (from 250 to 16,000), different length of the exemplars (from 1 to 50 time frames) and state-labeled versus word-labeled decoding.In addition, we compare the novel method for selecting exemplars, based on Sparse Classification, with a con ventional K-Nearest-Neighbor approach.For word-labeled decod ing we developed a Viterbi search that applies minim um and m axi m um duration constraints.It appears that Sparse Classification out performs KNN, while state-labeled decoding provides better per formance than word-labeled decoding.In all conditions the per formance increases with the size o f the collection.However, the optimal window length is 10 frames for state-labeled decoding, but 35 frames for word-labeled decoding.