Write a Blog >>
Thu 12 Nov 2020 01:05 - 01:06 at Virtual room 1 - Fairness

Recent works in DNN testing show that DNN based image classifiers are susceptible to confusion and bias errors. A DNN model, even robust trained model can be highly confused between certain pair of objects or highly bias towards some object than others. In this paper, we propose a differentiable distance metric, which is highly correlated with confusion errors. We propose a repairing approach by increasing the distance between two classes during retraining the model to reduce the confusion errors. We evaluate our approaches on both single-label and multi-label classification models and datasets. Our results show that our approach effectively reduce confusion errors with very slight accuracy reduce.

Thu 12 Nov
Times are displayed in time zone: (UTC) Coordinated Universal Time change

01:00 - 01:02
Talk
Research Papers
Sumon BiswasIowa State University, USA, Hridesh RajanIowa State University, USA
Link to publication DOI Pre-print Media Attached
01:03 - 01:04
Talk
Research Papers
Joymallya ChakrabortyNorth Carolina State University, USA, Suvodeep MajumderNorth Carolina State University, USA, Zhe YuNorth Carolina State University, USA, Tim MenziesNorth Carolina State University, USA
DOI
01:05 - 01:06
Talk
Student Research Competition
Yuchi TianColumbia University
DOI
01:07 - 01:08
Talk
Research Papers
Ye LiuNanyang Technological University, Singapore, Yi LiNanyang Technological University, Singapore, Shang-Wei LinNanyang Technological University, Singapore, Rong ZhaoNanyang Technological University, Singapore
DOI Pre-print
01:09 - 01:30
Talk
Paper Presentations
Joymallya ChakrabortyNorth Carolina State University, USA, Sumon BiswasIowa State University, USA, Ye LiuNanyang Technological University, Singapore, Yi LiNanyang Technological University, Singapore, Yuchi TianColumbia University, M: Christian BirdMicrosoft Research