You are on page 1of 4

Technical Writing & Presentation Skills

Analysis Of Research Paper

Group Members:
Abdul Mannan(2019-CS-37)
Taha Javed(2019-CS-50)
Title:
You Only Look Once: Unified, Real-Time Object Detection

Source:
http://pjreddie.com/yolo/

Authors

1. Joseph Redmon (Middlebury College and then University of Washington)

2. Santosh Divvala (Allen Institute for Artificial Intelligence)

3. Ross Girshick (University Of California, Facebook AI Research (FAIR))

4. Ali Farhadi (Professor at University Of Washington & Leader of research


team in Allen Institute for Artificial Intelligence)
Summary
The authors present a technique for real-time object class
detection with a Convolutional Neural Network (CNN). In
contrast to state-of-the-art systems such as R-CNN, the
proposed technique does not rely on region proposals and
instead detects objects with a single pass through the CNN. The
main contribution is the idea of dividing the image into a 7x7
grid, and having each cell predict a distribution over class
labels as well as a bounding-box for the object whose center
falls into it. As a result, the method is much faster than R-CNN,
as the remaining bottleneck of extracting region proposals has
been removed.

Research Design: Author uses observational design


Reason: He is comparing YOLO with other computer vision algorithms.
Abstract
We present YOLO, a new approach to object detection.
Prior work on object detection repurposes classifiers to perform
detection. Instead, we frame object detection as a regression
problem to spatially separated bounding boxes and
associated class probabilities. A single neural network predicts
bounding boxes and class probabilities directly from
full images in one evaluation. Since the whole detection
pipeline is a single network, it can be optimized end-to-end
directly on detection performance. Finally, YOLO learns very
general representations of objects. It outperforms other detection
methods, including DPM and R-CNN, when generalizing
from natural images to other domains like artwork.
Comparison And Contrast: In Introduction and separate comparison topic

Cause And Effect: In Conclusion

You might also like