ssdkeras
所属分类:数值算法/人工智能
开发工具:R
文件大小:14530KB
下载次数:0
上传日期:2017-12-05 08:59:07
上 传 者:
sh-1993
说明: ssdkeras,用Rstudio的Keras包装器实现单次多盒检测器
(ssdkeras,Implementing Single-Shot MultiBox Detector with Rstudio s Keras wrapper)
文件列表:
.Rbuildignore (59, 2017-12-05)
DESCRIPTION (491, 2017-12-05)
NAMESPACE (170, 2017-12-05)
R (0, 2017-12-05)
R\anchorboxlayer.R (15100, 2017-12-05)
R\keras_ssd7.R (13582, 2017-12-05)
R\ssd_batch_generator.R (19520, 2017-12-05)
R\ssd_box_encode_decode_utils.R (45003, 2017-12-05)
R\ssd_loss.R (9263, 2017-12-05)
lfw_faces_tagged.zip (14619552, 2017-12-05)
man (0, 2017-12-05)
man\build_model.Rd (532, 2017-12-05)
serena.jpg (232002, 2017-12-05)
train_ssd7_faces.R (10803, 2017-12-05)
val_02.png (171036, 2017-12-05)
ssdkeras package - Under Construction, use at your own risk
===========================================================
This is a R port to [Pierluigi Ferrari's Python implementation](https://github.com/pierluigiferrari/ssd_keras) of Single Shot Multibox Detector - [SSD](https://arxiv.org/abs/1512.02325).
See how to use it for detecting eyes, nose and mouth on the [LFW faces dataset](http://vis-www.cs.umass.edu/lfw/) in [this](http://giorasimchoni.com/2017/12/04/2017-12-04-snap-fu/) blog post.
Or get the gist with `train_ssd7_faces.R` with the data zipped at `lfw_faces_tagged.zip`.
You should get quite good results, e.g.:
![](val_02.png)
As said this is under construction though working at large. Only lightly tested on my Windows machine in which for example Python, Keras, Tensorflow are already installed. Other packages needed:
`tidyverse`, `stringr`, `reticulate`, `magick`
See `sessionInfo()`:
``` r
sessionInfo()
```
## Warning in readLines("~/sessionInfo_04122017.txt"): incomplete final line
## found on '~/sessionInfo_04122017.txt'
## R version 3.4.1 (2017-06-30)
## Platform: x86_***-w***-mingw32/x*** (***-bit)
## Running under: Windows 7 x*** (build 7601) Service Pack 1
##
## Matrix products: default
##
## locale:
## [1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United States.1252
## [3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
## [5] LC_TIME=English_United States.1252
##
## attached base packages:
## [1] stats graphics grDevices utils datasets methods base
##
## other attached packages:
## [1] bindrcpp_0.2 dplyr_0.7.2 purrr_0.2.3 readr_1.1.1 tidyr_0.6.3 tibble_1.3.3
## [7] ggplot2_2.2.1 tidyverse_1.1.1 magick_1.5 stringr_1.2.0 keras_2.0.9.9000
##
## loaded via a namespace (and not attached):
## [1] reticulate_1.3 reshape2_1.4.2 haven_1.1.0 lattice_0.20-35 colorspace_1.3-2
## [6] yaml_2.1.14 base***enc_0.1-3 rlang_0.1.1 foreign_0.8-69 glue_1.1.1
## [11] modelr_0.1.1 readxl_1.0.0 jpeg_0.1-8 bindr_0.1 plyr_1.8.4
## [16] tensorflow_1.4.1.9000 munsell_0.4.3 gtable_0.2.0 cellranger_1.1.0 rvest_0.3.2
## [21] psych_1.7.5 knitr_1.16 forcats_0.2.0 tfruns_1.0 parallel_3.4.1
## [26] broom_0.4.2 Rcpp_0.12.13 scales_0.5.0 jsonlite_1.5 mnormt_1.5-5
## [31] bmp_0.2 hms_0.3 png_0.1-7 stringi_1.1.5 animation_2.5
## [36] grid_3.4.1 tools_3.4.1 magrittr_1.5 lazyeval_0.2.0 whisker_0.3-2
## [41] readbitmap_0.1-4 pkgconfig_2.0.1 zeallot_0.0.6 xml2_1.1.1 lubridate_1.6.0
## [46] imager_0.40.2 assertthat_0.2.0 httr_1.2.1 R6_2.2.2 nlme_3.1-131
## [51] compiler_3.4.1
近期下载者:
相关文件:
收藏者: