Keyphrases
Weakly Supervised Object Detection
50%
Intersection over Union
50%
Regression Loss
50%
Weighting Factor
50%
Seat Belt Detection
50%
Seat Belt
50%
Bidirectional Long Short-term Memory (BiLSTM)
50%
Bounding Box
50%
Popular
37%
Mean Average Precision
36%
Context Module
33%
Region Proposal
33%
Object Parts
33%
Contextual Information
33%
State-of-the-art Techniques
26%
Bounding Box Regression
25%
Overlapping Region
25%
Surveillance Imaging
20%
Object Localization
16%
Image-level Annotations
16%
Mining Strategy
16%
PASCAL VOC
16%
Computer Vision
16%
Mining Method
16%
Object Boundary
16%
Multiple Instance Learning
16%
Localization Accuracy
16%
MS COCO
12%
Geometric Relationships
12%
Imbalance Problem
12%
COCO Dataset
12%
Convergence Rate
12%
Deep Learning Detection
12%
Object Detection Algorithm
12%
Loss Function
12%
Performance Gain
12%
High-risk Behavior
10%
Rigorous Comparisons
10%
Traffic Injury
10%
Passenger Safety
10%
Diagonal Systematic Sampling
10%
Bottom-left
10%
YOLOX
10%
Dynamic Integration
10%
Safety Enforcement
10%
Vehicle Occupants
10%
Outperformance
10%
Torso
10%
Second Best
10%
Traffic Surveillance
10%
Computer Science
Object Detection
100%
Mean Average Precision
50%
Overlapping Region
50%
Bidirectional Long Short-Term Memory Network
50%
Contextual Information
33%
Convergence Speed
25%
Deep Learning
25%
Performance Gain
25%
Deep Neural Network
16%
Attention (Machine Learning)
16%
Source Codes
16%
Experimental Result
16%
Computer Vision
16%
Object Boundary
16%
Annotation
16%
Vehicle Occupant
16%
Local Information
16%
Surrounding Region
16%
Instance Learning
16%
Multiple Instance
16%
