Algorithm

Pedestrian

Engineering

Transport engineering

Set (abstract data type)

Computer science

Pedestrian detection

Pattern recognition (psychology)

Artificial intelligence

Cluster analysis

Programming language

Computer vision

Optik

The existing real-time pedestrian detection method often loses part of the detection accuracy. For this reason, we proposed a real-time pedestrian detection algorithm based on tiny-yolov3. The proposed method uses K-means clustering on our training set to find the best priors. We improved the network structure of tiny-yolov3 to make it more accurate in pedestrian detection. From the experimental results, the proposed method has higher detection accuracy under the premise of satisfying real-time performance.

An improved tiny-yolov3 pedestrian detection algorithm