![](https://csdnimg.cn/release/download_crawler_static/11629714/bg6.jpg)
6
TABLE I
AP SCORES (%) ON THE MS COCO DATASET,AP
S
:AP OF SMALL
OBJECTS, AP
M
:AP OF MEDIUM OBJECTS, AP
L
:AP OF LARGE OBJECTS
Model AP
S
AP
M
AP
L
DSSD513 13.0 35.4 51.1
RetinaNet 24.1 44.2 51.2
G. YOLOv3
YOLOv3 [30] is an improved version of YOLOv2. First,
YOLOv3 uses multi-label classification (independent logistic
classifiers) to adapt to more complex datasets containing
many overlapping labels. Second, YOLOv3 uses three dif-
ferent scales feature maps to predict the bounding box. The
last convolutional layer predicts a 3-d tensor encoding class
predictions, objectness, and bounding box. Third, YOLOv3
proposes a deeper and robust feature extractor, called Darknet-
53, inspired by ResNet to get deeper.
According to results of experiments on MS COCO dataset,
YOLOv3 (AP:33%) performs on par with the SSD vari-
ant (DSSD513:AP:33.2%) on MS COCO metrics yet 3
times faster than it while quite a bit behind RetinaNet [31]
(AP:40.8%). But uses the old detection metric of mAP at
IOU= 0.5 (or AP
50
), YOLOv3 can achieve 57.9% mAP as
compared to DSSD513 of 53.3% and RetinaNet of 61.1%.
Due to the advantages of multi-scale predictions, YOLOv3 can
detect small objects even more but has comparatively worse
performance on medium and larger size objects.
H. RetinaNet
RetinaNet [31] is a one-stage object detector with focal loss
as classification loss function proposed by Tsung-Yi Lin et al.
[31] in February 2018. The architecture of RetinaNet is shown
in Fig.4 (c). R-CNN is a typical two-stage object detector. The
first stage generates a sparse set of region proposals and the
second stage classifies each candidate location. Owing to the
first stage filters out the majority of negative locations, two-
stage object detectors can achieve higher precision than one-
stage detectors which propose a dense set of candidate loca-
tions. The main reason is the extreme foreground-background
class imbalance when one-stage detectors train networks to
get convergence. So the authors proposed a loss function,
called focal loss, which can down-weight the loss assigned to
well-classified or easy examples, focusing on the hard training
examples and avoiding the vast number of easy negative ex-
amples overwhelming the detector during training. RetinaNet
inherits the fast speed of previous one-stage detectors while
greatly overcomes one-stage detectors difficult to training for
unbalanced positive and negative examples.
Experiments showed that RetinaNet with ResNet-101-FPN
backbone got 39.1% AP as compared to DSSD513 of 33.2%
AP on MS COCO test-dev dataset. With ResNeXt-101-FPN,
it made 40.8% AP far surpassing the-state-of-the-art one-
stage detector–DSSD513. RetinaNet improved the detection
precision on small and medium objects by a large margin.
I. SSD
SSD [8], a single-shot detector for multiple categories
within one-stage which directly predicting category scores
and box offsets for a fixed set of default bounding boxes of
different scales at each location in several feature maps with
different scales, as shown in Fig.4 (a). The default bounding
boxes have different aspect ratios and scales in each feature
map. In different feature maps, the scale of default bounding
boxes is computed with regularly space between the highest
layer and the lowest layer where each specific feature map
learns to be responsive to the particular scale of the objects.
For each default box, it predicts both the offsets and the
confidences for all object categories. Fig.3 (c) shows the
method. At training time, matching these default bounding
boxes to the ground truth boxes where the matched default
boxes as positive examples and the rest as negatives. For the
large amount of default boxes are negatives, the authors adopt
hard negative mining using the highest confidence loss for each
default box then picking the top ones to make the ratio between
the negatives and positives at most 3:1. As well, the authors
implement data augmentation which is proved an effective way
to enhance precision by a large margin.
Experiments showed that SSD512 had a competitive result
both mAP and speed with VGG-16 [24] backbone. SSD512
(input image size: 512 × 512) achieved mAP of 81.6% on
PASCAL VOC 2007 test set and 80.0% on PASCAL VOC
2012 test set as compared to Faster R-CNN (78.8%, 75.9%)
and YOLO (VOC2012: 57.9%). On MS COCO DET dataset,
SSD512 was better than Faster R-CNN in all criteria.
J. DSSD
DSSD [32] (Deconvolutional Single Shot Detector) is
a modified version of SSD (Single Shot Detector) which
adding prediction module and deconvolution module also
using ResNet-101 as backbone. The architecture of DSSD
is shown in Fig.4 (b). For prediction module, Fu et al. add
a residual block to each predicting layer, then do element-
wise addition of the outputs of prediction layer and residual
block. Deconvolution module is for increasing the resolution
of feature maps to strengthen features. Each deconvolution
layer followed by a prediction module is to predict a variety
of objects with different sizes. At training process, first pre-
training ResNet-101 based backbone network on the ILSVRC
CLS-LOC dataset, then using 321 × 321 inputs or 513 × 513
inputs training the original SSD model on detection dataset,
finally training the deconvolution module freezing all the
weights of SSD module. Experiments on both PASCAL VOC
dataset and MS COCO dataset show the effectiveness of
DSSD513 model, while the added prediction module and
deconvolution module bring 2.2% enhancement on PASCAL
VOC 2007 test dataset.
K. RefineDet
The whole network [33] contains two inter-connected mod-
ules, the anchor refinement module and the object detection
module. These two modules are connected by a transfer