%PDF-1.4
%
1 0 obj
<>
endobj
2 0 obj
<>
endobj
3 0 obj
<>stream
IEEE
2020 IEEE International Parallel and Distributed Processing Symposium Workshops (IPDPSW);2020; ; ;10.1109/IPDPSW50202.2020.00017
Deep neural network accelerator
weights compression
approximate deep neural network
accuracy/latency/energy trade-off
Improving Inference Latency and Energy of Network-on-Chip based Convolutional Neural Networks through Weights Compression
endstream
endobj
4 0 obj
<>stream
x+ |
endstream
endobj
5 0 obj
<>stream
xAO@WQRٶ[֘6kc.iwva/p0gIim*}εs/,mXyVdOhxw=s^'7`C/$?d=
endstream
endobj
6 0 obj
<>stream
x+ |
endstream
endobj
7 0 obj
<>stream
xAO@WQRٶ[֘6kc.iwva/p09gIim*}εs/,mXyVdOhxw=s^'7`C/$?e=
endstream
endobj
8 0 obj
<>stream
x+ |
endstream
endobj
9 0 obj
<>stream
xAO@WQSٶ[֘6kciwva/hX/4-J