Statements (61)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:cosmic_ray_detector
|
| gptkbp:application |
gptkb:robotics
Autonomous vehicles Augmented reality Surveillance systems |
| gptkbp:architecture |
gptkb:YOLO_(You_Only_Look_Once)
|
| gptkbp:coat_of_arms |
gptkb:17
|
| gptkbp:community_support |
Open-source community
|
| gptkbp:dependency |
gptkb:Python
gptkb:Open_CV gptkb:cu_DNN gptkb:CUDA gptkb:CMake |
| gptkbp:developed_by |
gptkb:Joseph_Redmon
|
| gptkbp:evaluates |
m AP (mean Average Precision)
Io U (Intersection over Union) FPS (Frames Per Second) |
| gptkbp:feature |
High speed
Batch normalization Skip connections Lightweight model Leaky Re LU activation Single-stage detector Anchor boxes Multi-class detection Bounding box regression Class probability prediction Feature pyramid network Low computational cost |
| https://www.w3.org/2000/01/rdf-schema#label |
Tiny YOLOv3-17l
|
| gptkbp:input_output |
RGB images
416x416 Bounding boxes with class labels |
| gptkbp:is_a_framework_for |
gptkb:Darknet
|
| gptkbp:is_compared_to |
Less accurate than YOLOv3
More efficient than YOLOv3 Less efficient than Tiny YOLOv4 More accurate than Tiny YOLOv2 |
| gptkbp:language |
gptkb:C
gptkb:C++ gptkb:Python |
| gptkbp:license |
GPL-3.0
|
| gptkbp:model |
Approximately 6 MB
Larger than Tiny YOLOv2 Small compared to YOLOv3 |
| gptkbp:performance |
Real-time object detection
Lower than full YOLOv3 |
| gptkbp:platform |
Cross-platform
|
| gptkbp:predecessor |
Tiny YOLOv2
|
| gptkbp:provides_information_on |
gptkb:COCO_dataset
Varies by application |
| gptkbp:release_date |
gptkb:2019
|
| gptkbp:repository |
gptkb:Git_Hub
|
| gptkbp:speed |
Faster than full YOLOv3
|
| gptkbp:successor |
Tiny YOLOv4
|
| gptkbp:training |
Transfer learning
Fine-tuning |
| gptkbp:tutorials |
Available online
|
| gptkbp:use_case |
Embedded systems
|
| gptkbp:bfsParent |
gptkb:Tiny_YOLOv3
|
| gptkbp:bfsLayer |
7
|