spatial-correlation-sampler 0.0.5

Correlation module for pytorch

Correlation module

this is a custom C++/Cuda implementation of Correlation module, used e.g. in FlowNetC

This tutorial was used as a basis for implementation.

• Build and Install C++ and CUDA extensions by executing python setup.py install,
• Benchmark C++ vs. CUDA by running python benchmark.py {cpu, cuda},
• Run gradient checks on the code by running python grad_check.py --backend {cpu, cuda}.

Requirements

This module is expected to compile for Pytorch 0.4.1, on Python > 3.5 and Python 2.7.

Installation

this module is available on pip

pip install spatial-correlation-sampler

Benchmark

• default parameters are from benchmark.py, FlowNetC parameters are same as use in FlowNetC with a batch size of 4, described in this paper, implemented here and here.
• Feel free to file an issue to add entries to this with your hardware !

CUDA Benchmark

• See here for a benchmark script working with NVIDIA's code, and Pytorch 0.3.
• Benchmark are launched with environment variable CUDA_LAUNCH_BLOCKING set to 1.
• Only float32 is benchmarked.

implementation	Correlation parameters	device	pass	min time	avg time
ours	default	980 GTX	forward	24.912 ms	25.202 ms
ours	default	980 GTX	backward	148.341 ms	148.827 ms
NVIDIA	default	980 GTX	forward	80.743 ms	81.534 ms
NVIDIA	default	980 GTX	backward	423.968 ms	425.396 ms
ours	FlowNetC	980 GTX	forward	6.888 ms	7.578 ms
ours	FlowNetC	980 GTX	backward	28.558 ms	28.984 ms
NVIDIA	FlowNetC	980 GTX	forward	8.640 ms	8.805 ms
NVIDIA	FlowNetC	980 GTX	backward	75.757 ms	76.873 ms

CPU Benchmark

• No other implementation is avalaible on CPU.

Correlation parameters	device	pass	min time	avg time
default	E5-2630 v3 @ 2.40GHz	forward	618.303 ms	626.618 ms
default	E5-2630 v3 @ 2.40GHz	backward	1052.563 ms	1083.407 ms
FlowNetC	E5-2630 v3 @ 2.40GHz	forward	339.769 ms	354.526 ms
FlowNetC	E5-2630 v3 @ 2.40GHz	backward	776.335 ms	785.781 ms