Skip to content

Latest commit

 

History

History
344 lines (234 loc) · 11.3 KB

GETTING_STARTED.md

File metadata and controls

344 lines (234 loc) · 11.3 KB
Raw

Getting Started

This page provides basic tutorials about the usage of MMDetection. For installation instructions, please see INSTALL.md.

Inference with pretrained models

We provide testing scripts to evaluate a whole dataset (COCO, PASCAL VOC, DOTA, HRSC2016 etc.), and also some high-level apis for easier integration to other projects.

Test a dataset

  • single GPU testing
  • multiple GPU testing
  • visualize detection results

You can use the following commands to test a dataset.

# single-gpu testing
python tools/test.py ${CONFIG_FILE} ${CHECKPOINT_FILE} [--out ${RESULT_FILE}] [--eval ${EVAL_METRICS}] [--show] [--data ${DATASET_NAME}]

# multi-gpu testing
./tools/dist_test.sh ${CONFIG_FILE} ${CHECKPOINT_FILE} ${GPU_NUM} [--out ${RESULT_FILE}] [--eval ${EVAL_METRICS}] [--data ${DATASET_NAME}]

Optional arguments:

  • RESULT_FILE: Filename of the output results in pickle format. If not specified, the results will not be saved to a file.
  • EVAL_METRICS: Items to be evaluated on the results. Allowed values are: proposal_fast, proposal, bbox, segm, keypoints.
  • --show: If specified, detection results will be plotted on the images and shown in a new window. It is only applicable to single GPU testing. Please make sure that GUI is available in your environment, otherwise you may encounter the error like cannot connect to X server.

Examples:

Assume that you have already downloaded the checkpoints to work_dirs/s2anet_r50_fpn_1x/.

  • Test S2A-Net on DOTA.
python tools/test.py configs/dota/s2anet_r50_fpn_1x.py \
    work_dirs/s2anet_r50_fpn_1x/epoch_12.pth \ 
    --out work_dirs/s2anet_r50_fpn_1x/res.pkl
  • Test S2A-Net on HRSC2016.
python tools/test.py configs/hrsc2016/s2anet_r50_fpn_3x_hrsc2016.py \
    work_dirs/s2anet_r50_fpn_3x_hrsc2016/epoch_36.pth \ 
    --out work_dirs/s2anet_r50_fpn_3x_hrsc2016/res.pkl \
    --data hrsc2016
  • Test S2A-Net with 4 GPUs.
./tools/dist_test.sh configs/dota/s2anet_r50_fpn_1x.py \
    work_dirs/s2anet_r50_fpn_1x/epoch_12.pth 4 \ 
    --out work_dirs/s2anet_r50_fpn_1x/res.pkl

*If you want to evaluate the result on DOTA test-dev, zip the files in work_dirs/s2anet_r50_fpn_1x/result_after_nms and submit it to the evaluation server.

Inference

To inference multiple images in a folder, you can run:

python demo/demo_inference.py ${CONFIG_FILE} ${CHECKPOINT} ${IMG_DIR} ${OUTPUT_DIR}

Train a model

MMDetection implements distributed training and non-distributed training, which uses MMDistributedDataParallel and MMDataParallel respectively.

All outputs (log files and checkpoints) will be saved to the working directory, which is specified by work_dir in the config file.

*Important*: The default learning rate in config files is for 8 GPUs and 2 img/gpu (batch size = 8*2 = 16). According to the Linear Scaling Rule, you need to set the learning rate proportional to the batch size if you use different GPUs or images per GPU, e.g., lr=0.01 for 4 GPUs * 2 img/gpu and lr=0.08 for 16 GPUs * 4 img/gpu.

Train with a single GPU

python tools/train.py ${CONFIG_FILE}

If you want to specify the working directory in the command, you can add an argument --work_dir ${YOUR_WORK_DIR}.

Train with multiple GPUs

./tools/dist_train.sh ${CONFIG_FILE} ${GPU_NUM} [optional arguments]

Optional arguments are:

  • --validate (strongly recommended): Perform evaluation at every k (default value is 1, which can be modified like this) epochs during the training.
  • --work_dir ${WORK_DIR}: Override the working directory specified in the config file.
  • --resume_from ${CHECKPOINT_FILE}: Resume from a previous checkpoint file.

Difference between resume_from and load_from: resume_from loads both the model weights and optimizer status, and the epoch is also inherited from the specified checkpoint. It is usually used for resuming the training process that is interrupted accidentally. load_from only loads the model weights and the training epoch starts from 0. It is usually used for finetuning.

Train with multiple machines

If you run MMDetection on a cluster managed with slurm, you can use the script slurm_train.sh.

./tools/slurm_train.sh ${PARTITION} ${JOB_NAME} ${CONFIG_FILE} ${WORK_DIR} [${GPUS}]

Here is an example of using 16 GPUs to train Mask R-CNN on the dev partition.

./tools/slurm_train.sh dev mask_r50_1x configs/mask_rcnn_r50_fpn_1x.py /nfs/xxxx/mask_rcnn_r50_fpn_1x 16

You can check slurm_train.sh for full arguments and environment variables.

If you have just multiple machines connected with ethernet, you can refer to pytorch launch utility. Usually it is slow if you do not have high speed networking like infiniband.

Useful tools

Analyze logs

You can plot loss/mAP curves given a training log file. Run pip install seaborn first to install the dependency.

python tools/analyze_logs.py plot_curve [--keys ${KEYS}] [--title ${TITLE}] [--legend ${LEGEND}] [--backend ${BACKEND}] [--style ${STYLE}] [--out ${OUT_FILE}]

Examples:

  • Plot the classification loss of some run.
python tools/analyze_logs.py plot_curve log.json --keys loss_cls --legend loss_cls
  • Plot the classification and regression loss of some run, and save the figure to a pdf.
python tools/analyze_logs.py plot_curve log.json --keys loss_cls loss_reg --out losses.pdf
  • Compare the bbox mAP of two runs in the same figure.
python tools/analyze_logs.py plot_curve log1.json log2.json --keys bbox_mAP --legend run1 run2

You can also compute the average training speed.

python tools/analyze_logs.py cal_train_time ${CONFIG_FILE} [--include-outliers]

The output is expected to be like the following.

-----Analyze train time of work_dirs/some_exp/20190611_192040.log.json-----
slowest epoch 11, average time is 1.2024
fastest epoch 1, average time is 1.1909
time std over epochs is 0.0028
average iter time: 1.1959 s/iter

Get the FLOPs and params (experimental)

We provide a script adapted from flops-counter.pytorch to compute the FLOPs and params of a given model.

python tools/get_flops.py ${CONFIG_FILE} [--shape ${INPUT_SHAPE}]

You will get the result like this.

==============================
Input shape: (3, 1280, 800)
Flops: 239.32 GMac
Params: 37.74 M
==============================

Note: This tool is still experimental and we do not guarantee that the number is correct. You may well use the result for simple comparisons, but double check it before you adopt it in technical reports or papers.

(1) FLOPs are related to the input shape while parameters are not. The default input shape is (1, 3, 1280, 800). (2) Some operators are not counted into FLOPs like GN and custom operators. You can add support for new operators by modifying mmdet/utils/flops_counter.py. (3) The FLOPs of two-stage detectors is dependent on the number of proposals.

Publish a model

Before you upload a model to AWS, you may want to (1) convert model weights to CPU tensors, (2) delete the optimizer states and (3) compute the hash of the checkpoint file and append the hash id to the filename.

python tools/publish_model.py ${INPUT_FILENAME} ${OUTPUT_FILENAME}

E.g.,

python tools/publish_model.py work_dirs/faster_rcnn/latest.pth faster_rcnn_r50_fpn_1x_20190801.pth

The final output filename will be faster_rcnn_r50_fpn_1x_20190801-{hash id}.pth.

Test the robustness of detectors

Please refer to ROBUSTNESS_BENCHMARKING.md.

How-to

Use my own datasets

The simplest way is to convert your dataset to existing dataset formats (COCO or PASCAL VOC).

Here we show an example of adding a custom dataset of 5 classes, assuming it is also in COCO format.

In mmdet/datasets/my_dataset.py:

from .coco import CocoDataset
from .registry import DATASETS


@DATASETS.register_module
class MyDataset(CocoDataset):

    CLASSES = ('a', 'b', 'c', 'd', 'e')

In mmdet/datasets/__init__.py:

from .my_dataset import MyDataset

Then you can use MyDataset in config files, with the same API as CocoDataset.

It is also fine if you do not want to convert the annotation format to COCO or PASCAL format. Actually, we define a simple annotation format and all existing datasets are processed to be compatible with it, either online or offline.

The annotation of a dataset is a list of dict, each dict corresponds to an image. There are 3 field filename (relative path), width, height for testing, and an additional field ann for training. ann is also a dict containing at least 2 fields: bboxes and labels, both of which are numpy arrays. Some datasets may provide annotations like crowd/difficult/ignored bboxes, we use bboxes_ignore and labels_ignore to cover them.

Here is an example.

[
    {
        'filename': 'a.jpg',
        'width': 1280,
        'height': 720,
        'ann': {
            'bboxes': <np.ndarray, float32> (n, 4),
            'labels': <np.ndarray, int64> (n, ),
            'bboxes_ignore': <np.ndarray, float32> (k, 4),
            'labels_ignore': <np.ndarray, int64> (k, ) (optional field)
        }
    },
    ...
]

There are two ways to work with custom datasets.

  • online conversion

    You can write a new Dataset class inherited from CustomDataset, and overwrite two methods load_annotations(self, ann_file) and get_ann_info(self, idx), like CocoDataset and VOCDataset.

  • offline conversion

    You can convert the annotation format to the expected format above and save it to a pickle or json file, like pascal_voc.py. Then you can simply use CustomDataset.

Develop new components

We basically categorize model components into 4 types.

  • backbone: usually an FCN network to extract feature maps, e.g., ResNet, MobileNet.
  • neck: the component between backbones and heads, e.g., FPN, PAFPN.
  • head: the component for specific tasks, e.g., bbox prediction and mask prediction.
  • roi extractor: the part for extracting RoI features from feature maps, e.g., RoI Align.

Here we show how to develop new components with an example of MobileNet.

  1. Create a new file mmdet/models/backbones/mobilenet.py.
import torch.nn as nn

from ..registry import BACKBONES


@BACKBONES.register_module
class MobileNet(nn.Module):

    def __init__(self, arg1, arg2):
        pass

    def forward(x):  # should return a tuple
        pass
    
    def init_weights(self, pretrained=None):
        pass
  1. Import the module in mmdet/models/backbones/__init__.py.
from .mobilenet import MobileNet
  1. Use it in your config file.
model = dict(
    ...
    backbone=dict(
        type='MobileNet',
        arg1=xxx,
        arg2=xxx),
    ...

For more information on how it works, you can refer to TECHNICAL_DETAILS.md (TODO).