Se ha denunciado esta presentación.
Utilizamos tu perfil de LinkedIn y tus datos de actividad para personalizar los anuncios y mostrarte publicidad más relevante. Puedes cambiar tus preferencias de publicidad en cualquier momento.
Video Object Detection
Challenge
Box-level post-processing
*Feature level learning
• Flow-Guided Feature Aggregation for Video Object Detection
• Deep Feat...
Flow-Guided Feature Aggregation
1. Warping:
2. Aggregation:
3. Adaptive weight:
Flow-Guided Feature Aggregation Deep Feature Flow
Flow-Guided Feature Aggregation Deep Feature Flow
Towards High Performance Video Object Detection
Fully Motion-Aware Network
Pixel level feature calibration is bad in:
• Appearance dynamic changes
• Occlusion
ROI-level m...
Fully Motion-Aware Network
Paper Date Base detector Backbone Tracking? Optical flow? Online? mAP(%) FPS (Titan X)
Seq-NMS Feb 2016 R-FCN ResNet101 no...
Próxima SlideShare
Cargando en…5
×

Video object detection

63 visualizaciones

Publicado el

Qi Rao

Publicado en: Tecnología
  • Sé el primero en comentar

  • Sé el primero en recomendar esto

Video object detection

  1. 1. Video Object Detection
  2. 2. Challenge
  3. 3. Box-level post-processing *Feature level learning • Flow-Guided Feature Aggregation for Video Object Detection • Deep Feature Flow for Video Recognition • Towards High Performance Video Object Detection • Fully Motion-Aware Network for Video Object Detection
  4. 4. Flow-Guided Feature Aggregation 1. Warping: 2. Aggregation: 3. Adaptive weight:
  5. 5. Flow-Guided Feature Aggregation Deep Feature Flow
  6. 6. Flow-Guided Feature Aggregation Deep Feature Flow
  7. 7. Towards High Performance Video Object Detection
  8. 8. Fully Motion-Aware Network Pixel level feature calibration is bad in: • Appearance dynamic changes • Occlusion ROI-level motion: Motion guided calibration: Aggregation:
  9. 9. Fully Motion-Aware Network
  10. 10. Paper Date Base detector Backbone Tracking? Optical flow? Online? mAP(%) FPS (Titan X) Seq-NMS Feb 2016 R-FCN ResNet101 no no no 76.8 2.3 T-CNN Apr 2016 RCNN DeepIDNet+CRAF T yes no no 73.8 - DFF Nov 2016 R-FCN ResNet101 no yes yes 73.0 29 TPN Feb 2017 TPN GoogLeNet yes no no 68.4 - FGFA Mar 2017 R-FCN ResNet101 no yes yes 76.3 1.4 FGFA + Seq-NMS 29 Mar 2017 R-FCN ResNet101 no yes no 78.4 1.14 D&T Oct 2017 R-FCN(15 anchors) ResNet101 yes no no 79.8 7.09 STMN Dec 2017 R-FCN ResNet101 no no no 80.5 - Scale-time-lattice 16 Apr 2018 Faster RCNN(15 anchors) ResNet101 no no no 79.6 20 Scale-time-lattice Apr 2018 Faster RCNN(15 anchors) ResNet101 no no no 79.0 62 SSN (per-frame baseline for STSN) Mar 2018 R-FCN Deformable ResNet101 no no yes 76.0 - STSN Mar 2018 R-FCN Deformable ResNet101 no no yes 78.9 - STSN+Seq-NMS Mar 2018 R-FCN Deformable ResNet101 no no no 80.4 - MANet Sep. 2018 R-FCN ResNet101 no yes yes 78.1 5 MANet+Seq-NMS Sep. 2018 R-FCN ResNet101 no yes no 80.3 - Tracklet- Conditioned Detection Nov 2018 R-FCN ResNet101 yes no yes 78.1 - Tracklet- Conditioned Detection+DCNv2 Nov 2018 R-FCN ResNet101 yes no yes 82.0 -

×