Fixes in opts

README.md

@@ -12,13 +12,13 @@ The results are available on the [project website](http://www.di.ens.fr/willow/r
 This is a joint work of [Vadim Kantorov](http://vadimkantorov.com), [Maxime Oquab](http://github.com/qassemoquab), [Minsu Cho](http://www.di.ens.fr/~mcho), and [Ivan Laptev](http://www.di.ens.fr/~laptev).
 
 # Running the code
-1. Install the dependencies: [Torch](http://github.com/torch/distro) with [cuDNN](http://developer.nvidia.com/cudnn) support; [HDF5](http://www.hdfgroup.org/HDF5/); [matio](http://github.com/tbeu/matio); [protobuf](http://github.com/google/protobuf); Luarocks packages [rapidjson](http://github.com/xpol/lua-rapidjson), [hdf5](http://github.com/deepmind/torch-hdf5), [matio](http://github.com/soumith/matio-ffi.torch), [loadcaffe](http://github.com/szagoruyko/loadcaffe).
+1. Install the dependencies: [Torch](http://github.com/torch/distro) with [cuDNN](http://developer.nvidia.com/cudnn) support; [HDF5](http://www.hdfgroup.org/HDF5/); [matio](http://github.com/tbeu/matio); [protobuf](http://github.com/google/protobuf); Luarocks packages [rapidjson](http://github.com/xpol/lua-rapidjson), [hdf5](http://github.com/deepmind/torch-hdf5), [matio](http://github.com/soumith/matio-ffi.torch), [loadcaffe](http://github.com/szagoruyko/loadcaffe), [xml](https://://github.com/lubyk/xml).
 
   We strongly recommend using [wigwam](http://wigwam.in/) for this (fix the paths to `nvcc` and `libcudnn.so` before running the command):
 
   ```
   $ wigwam install torch hdf5 matio protobuf -DPATH_TO_NVCC="/path/to/cuda/bin/nvcc" -DPATH_TO_CUDNN_SO="/path/to/cudnn/lib64/libcudnn.so"
-  $ wigwam install lua-rapidjson lua-hdf5 lua-matio lua-loadcaffe
+  $ wigwam install lua-rapidjson lua-hdf5 lua-matio lua-loadcaffe lua-xml
   ```
 2. Clone this repository, change the current directory to `contextlocnet`, and compile the ROI pooling module:
 
@@ -27,30 +27,30 @@ This is a joint work of [Vadim Kantorov](http://vadimkantorov.com), [Maxime Oqua
   $ cd contextlocnet
   $ (cd ./model && luarocks make)
   ```
-3. Download the [VOC 2007](http://host.robots.ox.ac.uk/pascal/VOC/voc2007/) dataset and Koen van de Sande's [selective search windows](http://koen.me/research/selectivesearch/) for VOC 2007 and the [VGG-F](https://gist.github.com/ksimonyan/a32c9063ec8e1118221a) model. Optionally download the [VOC 2012](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/) and Rob Girshick's [selective search windows](https://github.com/rbgirshick/fast-rcnn/blob/master/data/scripts/fetch_fast_rcnn_models.sh) by manually downloading [VOC 2012 test data tarball](http://host.robots.ox.ac.uk:8080/eval/downloads/VOC2012test.tar) to `data/common` and then running the second command):
+3. Download the [VOC 2007](http://host.robots.ox.ac.uk/pascal/VOC/voc2007/) dataset and Koen van de Sande's [selective search windows](http://koen.me/research/selectivesearch/) for VOC 2007 and the [VGG-F](https://gist.github.com/ksimonyan/a32c9063ec8e1118221a) model by running the first command. Optionally download the [VOC 2012](http://host.robots.ox.ac.uk/pascal/VOC/voc2012/) and Rob Girshick's [selective search windows](https://github.com/rbgirshick/fast-rcnn/blob/master/data/scripts/fetch_fast_rcnn_models.sh) by manually downloading [VOC 2012 test data tarball](http://host.robots.ox.ac.uk:8080/eval/downloads/VOC2012test.tar) to `data/common` and then running the second command:
 
   ```
   $ make -f data/common/Makefile download_and_extract_VOC2007 download_VGGF
   $ # make -f data/common/Makefile download_and_extract_VOC2012
   ```
-4. Preprocess the VOC data and convert the VGG-F model to Torch format:
+4. Choose a dataset, preprocess it, and convert the VGG-F model to the Torch format:
 
   ```
+  $ export DATASET=VOC2007
   $ th preprocess.lua VOC VGGF
   ```
-5. Choose a dataset, a model (our best model is `model/contrastive_s.lua`, other choices are `model/contrastive_a.lua`, `model/additive.lua`, and `model/wsddn_repro.lua`) and train it:
+5. Select a GPU and train a model  model (our best model is `model/contrastive_s.lua`, other choices are `model/contrastive_a.lua`, `model/additive.lua`, and `model/wsddn_repro.lua`):
 
   ```
-  $ export MODEL=model/contrastive_s.lua
-  $ export DATASET=VOC2007
-  $ th train.lua
+  $ export CUDA_VISIBLE_DEVICES=0
+  $ th train.lua model/contrastive_s.lua # will produce ./data/model_epoch30.h5 and ./data/log.json
   ```
 6. Test the trained model and compute CorLoc and mAP:
 
   ```
-  $ SUBSET=trainval th test.lua
+  $ SUBSET=trainval th test.lua ./data/model_epoch30.h5
   $ th corloc.lua
-  $ SUBSET=test th test.lua
+  $ SUBSET=test th test.lua ./data/model_epoch30.h5
   $ th detection_mAP.lua
   ```
 

opts.lua

@@ -43,17 +43,24 @@ PATHS =
 	DETECTION_MAP = paths.concat(DATA, 'detection_map.json'),
 }
 
-local BASE_MODEL = 'VGGF'
 local DATASET = os.getenv('DATASET') or 'VOC2007'
-local ROI_FACTOR = tonumber(os.getenv('ROI_FACTOR')) or 1.8
+local NUM_EPOCHS = tonumber(os.getenv('NUM_EPOCHS')) or 30
+local SUBSET = os.getenv('SUBSET') or 'trainval'
+local BASE_MODEL = 'VGGF'
 
 opts = {
 	DATASET = DATASET,
 	BASE_MODEL = BASE_MODEL,
-	ROI_FACTOR = ROI_FACTOR,
+	ROI_FACTOR = 1.8,
 	OUTPUTS_FIELDS = {'output_prod'},
+	SEED = 2,
+	NUM_EPOCHS = NUM_EPOCHS,
+	NUM_SCALES = 5,
+	SUBSET = SUBSET,
 	PATHS = 
 	{
+		MODEL = arg[1],
+
 		DATA = DATA,
 		DATA_COMMON = DATA_COMMON,
 

preprocess.lua

@@ -75,6 +75,7 @@ function VOC()
 end
 
 for _, a in ipairs(arg) do
-	print('Preprocessing', a, '...')
+	print('Preprocessing', a)
 	_G[a]()
 end
+print('Done')

test.lua

@@ -6,11 +6,8 @@ dofile('perf/util.lua')
 
 assert(os.getenv('CUDA_VISIBLE_DEVICES') ~= nil and cutorch.getDeviceCount() <= 1, 'SHOULD RUN ON ONE GPU FOR NOW')
 
-opts.NUM_SCALES = tonumber(os.getenv('NUM_SCALES')) or 5
-opts.SUBSET = os.getenv('SUBSET') or 'trainval'
-opts.MODEL = arg[1]
 
-loaded = model_load(opts.MODEL, opts)
+loaded = model_load(opts.PATHS.MODEL, opts)
 
 meta = {
 	opts = opts,

train.lua

@@ -2,17 +2,12 @@ dofile('opts.lua')
 dofile('util.lua')
 dofile('dataset.lua')
 dofile('model/util.lua')
-dofile('perf/util.lua')
 
 require 'optim'
 dofile('fbnn_Optim.lua')
 
 assert(os.getenv('CUDA_VISIBLE_DEVICES') ~= nil and cutorch.getDeviceCount() <= 1, 'SHOULD RUN ON ONE GPU FOR NOW')
 
-opts.SEED = tonumber(os.getenv('SEED')) or torch.initialSeed()
-opts.NUM_EPOCHS = tonumber(os.getenv('NUM_EPOCHS')) or 30
-opts.MODEL = arg[1]
-
 torch.manualSeed(opts.SEED)
 cutorch.manualSeedAll(opts.SEED)
 
@@ -31,16 +26,16 @@ example_loader_options_preset = {
 	}
 }
 
-if paths.extname(opts.MODEL) == 'lua' then
-	loaded = model_load(opts.MODEL, opts)
+if paths.extname(opts.PATHS.MODEL) == 'lua' then
+	loaded = model_load(opts.PATHS.MODEL, opts)
 	meta = {
 		model_path = loaded.model_path,
 		opts = opts,
 		example_loader_options = example_loader_options_preset
 	}
 	log = {{meta = meta}}
 else
-	loaded = model_load(opts.MODEL)
+	loaded = model_load(opts.PATHS.MODEL)
 	meta = loaded.meta
 	log = loaded.log
 	previous_epoch = loaded.epoch