Skip to content

Latest commit

 

History

History
111 lines (91 loc) · 6.31 KB

FLAGS.md

File metadata and controls

111 lines (91 loc) · 6.31 KB
Raw

Flags

This file documents all available flags for all scripts.

Common flags

There are some common flags that are shared by many Lua scripts:

  • -checkpoint: Path to the checkpoint containing the model to use
  • -image_size: Before being processed by the model, images will be resized so their max side length is equal to this; larger values will make the model run more slowly.
  • -test_rpn_nms_thresh: Threshold for non-maximum suppression among region proposals in the region proposal network; should be a number between 0 and 1.
  • -test_final_nms_thresh: Threshold for non-maximum suppression among final output boxes; should be a number between 0 and 1. Using a larger value will cause the model to produce more detections, but they will have more overlap with each other.
  • -num_proposals: The number of region proposals that should be run through the recognition network and language model. Using a larger number will tend to give better results, but will also make the model run slower.
  • -gpu: The (zero-indexed) index of the GPU to use. Setting this to -1 will run in CPU-only mode.
  • -use_cudnn: Setting this to 1 will enable the use of the cuDNN library when running in GPU mode. Setting this to 0 will fall back to the default torch implementation of cuDNN layers.
  • -anchor_type: The type of anchors used. Could choose from densecap, voc and coco (voc and coco is that in the faster-rcnn code)

preprocess2.py

The script preprocess2.py is used to preprocess a dataset of images, regions from Pascal VOC, and convert the entire dataset (except raw images) to an HDF5 file and a JSON file to be read by Lua scripts.

During preprocessing we remove images with no region annotations.

The following flags are available:

Input data:

  • --data_dir: Directory of the VOC data, could exactly be what you downloaded from VOC website. You can also merge the images and annotations to another directory which has the same structure of the normal VOC directory.

Output data:

  • --json_output: Path where the output JSON file should be written; this file contains the mapping from class label indices to strings.
  • --h5_output: Path where the output HDF5 file should be written; this file contains all regions, class labels and etc.

Options:

  • --train_split: Image ids of training data. (Default is the public training splitin the dataset)
  • --val_split: like above
  • --test_split: like above

** You can also use preprocess.py which will contain images in the dataset, in that way, you need to use 'DataLoader.lua' instead of 'DataLoader_new.lua' **

run_model.lua

The script run_model.lua is used to run a trained model. It can take as input a single image, a directory of images, or an HDF5 file produced by preprocess.py. It can output JSON files to be viewed by the HTML visualizer, or image files with a prespecified number of detections baked in.

The following flags are available:

Model options:

Input options: At least one of the following must be provided:

  • -input_image: Path to a single image on which to run the model
  • -input_dir: Path to a directory of images on which to run the model
  • -input_split: One of train, val, or test indicating a split of the Visual Genome data on which to run
    • -splits_json: Path to a JSON file containing splits of the Visual Genome data; only used if -input_split is provided.
    • -vg_img_root_dir: Path to a directory containing all of the Visual Genome images; only used if -input_split is provided.

Output options: The script can produce two types of outputs. By default it will write images and JSON files to the vis/data directory, which can be visualized with the HTML viewer. By providing the -output_dir flag the script will also produce images with a certain number of detections and captions baked in.

  • -max_images: The maximum number of images to process
  • -output_dir: If provided, the directory to which images with "baked in" detections should be written
    • '-num_to_draw: The number of detections to draw in output images; only used if -output_dir` is provided
    • -text_size: The font size to use for captions in output images; only used if -output_dir is provided
    • -box_width: Width (in pixels) of boxes in output images; only used if -output_dir is provided.
  • -output_vis: If 1 (default) then output JSON files and images for the HTML viewer -output_vis_dir: Directory to which images and JSON files (for the HTML viewer) should be written; default is vis/data.

train.lua

The script train.lua uses the HDF5 and JSON files produced by preprocess2.py to train models. This script periodically checks accuracy on the validation set, and saves checkpoints containing trained models.

In most cases the only settings you will need to change are -learning_rate, -checkpoint_start_from to start training from a checkpoint rather than training from scratch, -finetune_cnn_after to enable finetuning of the CNN, and -checkpoint_path to change the location of saved checkpoints.

~~

evaluate_model.lua

The script evaluate_model.lua is used to quantitatively evaluate a trained model using our mean average precision metric; a model is run on images in a preprocessed dataset that has been created by preprocess.py. This script assumes that METEOR has been downloaded by running scripts/setup_eval.sh, and requires both Python and Java. ~~

The following flags are available:

Model options:

Other options:

  • -data_h5: Path to an HDF5 file created by preprocess.py.
  • -data_json: Path to a JSON file created by preprocess.py.
  • -split: Which split to use; either train, val, or test.
  • -max_images: How many images to use for evaluation; set to -1 to use the entire split.