Alert button

PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track

Add code
Bookmark button
Alert button
Sep 04, 2018
Takuya Akiba, Tommi Kerola, Yusuke Niitani, Toru Ogawa, Shotaro Sano, Shuji Suzuki

Figure 1 for PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track
Figure 2 for PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track
Figure 3 for PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track
Figure 4 for PFDet: 2nd Place Solution to Open Images Challenge 2018 Object Detection Track

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon

We present a large-scale object detection system by team PFDet. Our system enables training with huge datasets using 512 GPUs, handles sparsely verified classes, and massive class imbalance. Using our method, we achieved 2nd place in the Google AI Open Images Object Detection Track 2018 on Kaggle.

* Technical report for Open Images Challenge 2018 Object Detection Track  
View paper onarxiv icon

Share this with someone who'll enjoy it:

Twitter IconFacebook IconLinkedin IconWhatsapp IconMessenger Icon