add model path searcher and remove hardwritten file paths from benchmark configs; pack ppresnet, mobilenet & crnn with labels; fix palm det data; add flags to enable models of different precision separately (#123)

* add backend TIMVX & target NPU

* update benchmarking results on Khadas VIM3 NPU

* fix wrong column header

* update readme

* re-order column KV3-NPU

* re-order KV3-NPU specs line

* add additional description regarding TIM-VX backend and NPU target for OpenCV DNN

* impl wrapper & demo

* add data for object tracking benchmark

* impl benchmark for DaSiamRPN

* update benchmark results for DaSiamRPN

* impl dataloader factory

* set a type to init all instances for benchmarks

* create metrics for different types of behavior

* workable impl calling utils.METRICS in benchmark

* add PPResNet impl and demo

* add benchmark impl and results for PPResNet

* SFace and demo impl

* benchmark support and results for SFace

* Improve benchmark configurations:

* Move data downloading from configs to download_data.py. Add an alternative download link.

* Add Data class to operate data loading and indexing.

* Add Metric class to operate benchmark runs.

* Benchmark results are now the median or geometric mean of benchmark
  runs.

* benchmark framework: benchmarks based on configs

* added impl and benchmark for YuNet (face detection)

* added impl and benchmark for DB (text detection)

* added impl and benchmark for CRNN (text recognition)