INIS
matrices
100%
power
100%
architecture
100%
vectors
75%
size
50%
units
50%
reduction
50%
kernels
50%
applications
25%
design
25%
comparative evaluations
25%
weight
25%
cost
25%
learning
25%
speed
25%
demand
25%
accuracy
25%
compression
25%
optimization
25%
speech
25%
tables
25%
Computer Science
Vector Multiplication
100%
Long Short-Term Memory Networks
100%
Very large-scale integration (VLSI) architecture
100%
World Application
33%
Field Programmable Gate Arrays
33%
Memory Architecture
33%
Power Efficient
33%
Reduce Complexity
33%
Pipelining
33%
Long Short-Term Memory Network
33%
Speech Recognition
33%
Sequential Learning
33%
Fixed Points
33%
Complexity Reduction
33%
Engineering
Circulant
100%
Long Short-Term Memory
100%
Kernel Size
33%
Lookup Table
16%
Field Programmable Gate Arrays
16%
Complexity Reduction
16%
Real World Application
16%
Learning Task
16%
Compression Network
16%
Flip Flop Circuits
16%
Fixed Points
16%
Neuroscience
Short-Term Memory
100%
Speech Recognition
33%