Block-Based High Performance CNN Architectures for Frame-Level Overlapping Speech Detection.
Midia YousefiJohn H. L. HansenPublished in: IEEE ACM Trans. Audio Speech Lang. Process. (2021)
Keyphrases
- automatic detection
- object detection
- detection method
- speech recognition
- detection rate
- cellular neural networks
- noisy environments
- higher level
- detection accuracy
- frame rate
- audio visual
- spoken language
- neural network
- coarse grained
- levels of abstraction
- false alarms
- change detection
- video frames
- anomaly detection
- face recognition
- image segmentation
- social networks