TY - RPRT TI - Decomposed Temporal Dynamic CNN: Efficient Time-Adaptive Network for Text-Independent Speaker Verification Explained with Speaker Activation Map AU - Seong-Hu Kim AU - Hyeonuk Nam AU - Yong-Hwa Park PY - 2022 UR - https://arxiv.org/abs/2203.15277 ID - 2203.15277 ER -