Learning Data Augmentation Strategies for Object Detection

Data augmentation is a critical component of training deep learning models. Although data augmentation has been shown to significantly improve image classification, its potential has not been thoroughly investigated for object detection... (read more)

PDF Abstract ECCV 2020 PDF ECCV 2020 Abstract

Results from the Paper


TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK USES EXTRA
TRAINING DATA
RESULT BENCHMARK
Object Detection COCO test-dev NAS-FPN (AmoebaNet-D, learned aug) box AP 50.7 # 29
APS 34.2 # 18
APM 55.5 # 19
APL 64.5 # 19
Object Detection PASCAL VOC 2007 Faster R-CNN (ResNet-101, learned aug) MAP 78.7% # 16

Methods used in the Paper


METHOD TYPE
Sigmoid Activation
Activation Functions
Tanh Activation
Activation Functions
Entropy Regularization
Regularization
Residual Connection
Skip Connections
PPO
Policy Gradient Methods
1x1 Convolution
Convolutions
LSTM
Recurrent Neural Networks
Batch Normalization
Normalization
ReLU
Activation Functions
NAS-FPN
Feature Extractors
SGD
Stochastic Optimization
Step Decay
Learning Rate Schedules
Image Scale Augmentation
Image Data Augmentation
Random Horizontal Flip
Image Data Augmentation
Focal Loss
Loss Functions
FPN
Feature Extractors
RetinaNet
Object Detection Models
Spatially Separable Convolution
Convolutions
Average Pooling
Pooling Operations
AmoebaNet
Convolutional Neural Networks
RPN
Region Proposal
Softmax
Output Functions
RoIPool
RoI Feature Extractors
Faster R-CNN
Object Detection Models
Bottleneck Residual Block
Skip Connection Blocks
Global Average Pooling
Pooling Operations
Residual Block
Skip Connection Blocks
Kaiming Initialization
Initialization
Max Pooling
Pooling Operations
Convolution
Convolutions
ResNet
Convolutional Neural Networks