Hello,

I found that the 'anchor_scale' is being used in faster_rcnn config files, and have tried to see what it does exactly. Whenever I tried to change this from [8] to some other values or multiple values like [8, 16, 32] to get more anchors (as I know, # of anchor = len(anchor_scale) * len(anchor_ratio)), I could see the training accuracy fluctuates a lot.

and also, I could see mAP of validation was 0 for all class. Is it related to anchor scale?

Config file I am using looks like below.

##Config

model settings

model = dict( type='FasterRCNN', pretrained='torchvision://resnet50', backbone=dict( type='ResNet', depth=50, num_stages=4,

out_indices=(3,), # more direct

    out_indices=(0, 1, 2, 3), # more general, handle at backbone (resnet.py) with conditions
    frozen_stages=1,
    style='pytorch',
    fpn_extension=False),
rpn_head=dict(
    type='RPNHead',
    in_channels=256,
    feat_channels=256,
    anchor_scales=[8], 
    anchor_ratios=[0.5, 1.0, 2.0],
    anchor_strides=[16],
    target_means=[.0, .0, .0, .0],
    target_stds=[1.0, 1.0, 1.0, 1.0],
    loss_cls=dict(
        type='CrossEntropyLoss', use_sigmoid=True, loss_weight=1.0),
    loss_bbox=dict(type='SmoothL1Loss', beta=1.0 / 9.0, loss_weight=1.0)),
bbox_roi_extractor=dict(
    type='SingleRoIExtractor',
    roi_layer=dict(type='RoIAlign', out_size=7, sample_num=2),
    out_channels=256,
    featmap_strides=[4, 8, 16, 32]),
bbox_head=dict(
    type='SharedFCBBoxHead',
    num_fcs=2,
    in_channels=256,
    fc_out_channels=1024,
    roi_feat_size=7,
    num_classes=21,
    target_means=[0., 0., 0., 0.],
    target_stds=[0.1, 0.1, 0.2, 0.2],
    reg_class_agnostic=False,
    loss_cls=dict(
        type='CrossEntropyLoss', use_sigmoid=False, loss_weight=1.0),
    loss_bbox=dict(type='SmoothL1Loss', beta=1.0, loss_weight=1.0)))

model training and testing settings

train_cfg = dict( rpn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.7, neg_iou_thr=0.3, min_pos_iou=0.3, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=256, pos_fraction=0.5, neg_pos_ub=-1, add_gt_as_proposals=False), allowed_border=0, pos_weight=-1, debug=False), rpn_proposal=dict( nms_across_levels=False, nms_pre=2000, nms_post=2000, max_num=2000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( assigner=dict( type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, ignore_iof_thr=-1), sampler=dict( type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=-1, add_gt_as_proposals=True), pos_weight=-1, debug=False)) test_cfg = dict( rpn=dict( nms_across_levels=False, nms_pre=1000, nms_post=1000, max_num=1000, nms_thr=0.7, min_bbox_size=0), rcnn=dict( score_thr=0.05, nms=dict(type='nms', iou_thr=0.5), max_per_img=100)

soft-nms is also supported for rcnn testing

# e.g., nms=dict(type='soft_nms', iou_thr=0.5, min_score=0.05)

)

dataset settings

dataset_type = 'VOCDataset'

data_root = 'data/VOCdevkit/'

data_root = '/shared/data/Pascal_VOC/VOCdevkit/' img_norm_cfg = dict( mean=[123.675, 116.28, 103.53], std=[58.395, 57.12, 57.375], to_rgb=True) train_pipeline = [ dict(type='LoadImageFromFile'), dict(type='LoadAnnotations', with_bbox=True), dict(type='Resize', img_scale=(1000, 600), keep_ratio=True), dict(type='RandomFlip', flip_ratio=0.5), dict(type='Normalize', img_norm_cfg), dict(type='Pad', size_divisor=32), dict(type='DefaultFormatBundle'), dict(type='Collect', keys=['img', 'gt_bboxes', 'gt_labels']), ] test_pipeline = [ dict(type='LoadImageFromFile'), dict( type='MultiScaleFlipAug', img_scale=(1000, 600), flip=False, transforms=[ dict(type='Resize', keep_ratio=True), dict(type='RandomFlip'), dict(type='Normalize', img_norm_cfg), dict(type='Pad', size_divisor=32), dict(type='ImageToTensor', keys=['img']), dict(type='Collect', keys=['img']), ]) ] data = dict( imgs_per_gpu=8, workers_per_gpu=2, train=dict( type='RepeatDataset', times=3, dataset=dict( type=dataset_type, ann_file=[ data_root + 'VOC2007/ImageSets/Main/trainval.txt', data_root + 'VOC2012/ImageSets/Main/trainval.txt' ], img_prefix=[data_root + 'VOC2007/', data_root + 'VOC2012/'], pipeline=train_pipeline)), val=dict( type=dataset_type, ann_file=data_root + 'VOC2007/ImageSets/Main/test.txt', img_prefix=data_root + 'VOC2007/', pipeline=test_pipeline), test=dict( type=dataset_type, ann_file=data_root + 'VOC2007/ImageSets/Main/test.txt', img_prefix=data_root + 'VOC2007/', pipeline=test_pipeline)) evaluation = dict(interval=1, metric='mAP')

optimizer

optimizer = dict(type='SGD', lr=0.01, momentum=0.9, weight_decay=0.0001) optimizer_config = dict(grad_clip=dict(max_norm=35, norm_type=2))

learning policy

lr_config = dict(policy='step', step=[3]) # actual epoch = 3 * 3 = 9 checkpoint_config = dict(interval=1)

yapf:disable

log_config = dict( interval=50, hooks=[ dict(type='TextLoggerHook'),

dict(type='TensorboardLoggerHook')

])

yapf:enable

runtime settings

total_epochs = 4 # actual epoch = 4 * 3 = 12 dist_params = dict(backend='nccl') log_level = 'INFO' work_dir = './work_dirs/study1_train1' load_from = None resume_from = None workflow = [('train', 1)]

workflow = [('train', 1), ('val', 1)]

Thank you.

open-mmlab / mmdetection

what actually anchor_scale does? #2124