lamply / dan-caffe Goto Github PK

A Caffe implementations of Deep Alignment Network.

• Python custom layer: training and inference
• C++ custom layer: inference in CPU

This project is an archived project, which has been stopped maintaining after 2018 and is for reference only.

Paper: Stage 1: Normalization is set to: centers Failure threshold is set to: 0.08 Processing common subset of the 300W public test set (test sets of LFPW and HELEN) Average error: 0.0493735770262 Processing challenging subset of the 300W public test set (IBUG dataset) Average error: 0.0896375119228 Showing results for the entire 300W pulic test set (IBUG dataset, test sets of LFPW and HELEN Average error: 0.0572627369841 AUC @ 0.08: 0.331357039187 Failure rate: 0.161103047896

Stage 2: Normalization is set to: centers Failure threshold is set to: 0.08 Processing common subset of the 300W public test set (test sets of LFPW and HELEN) Average error: 0.0441422118048 Processing challenging subset of the 300W public test set (IBUG dataset) Average error: 0.0755380425785 Showing results for the entire 300W pulic test set (IBUG dataset, test sets of LFPW and HELEN Average error: 0.0502937896778 AUC @ 0.08: 0.390457789066 Failure rate: 0.0812772133527

My: Stage 1: Normalization is set to: centers Failure threshold is set to: 0.08 Processing common subset of the 300W public test set (test sets of LFPW and HELEN) Average error: 0.0528969781986 Processing challenging subset of the 300W public test set (IBUG dataset) Average error: 0.0936767290662 Showing results for the entire 300W pulic test set (IBUG dataset, test sets of LFPW and HELEN Average error: 0.0608872051465 AUC @ 0.08: 0.293641146589 Failure rate: 0.162554426705

Stage 2: Normalization is set to: centers Failure threshold is set to: 0.08 Processing common subset of the 300W public test set (test sets of LFPW and HELEN) Average error: 0.0430267915831 Processing challenging subset of the 300W public test set (IBUG dataset) Average error: 0.0795358445265 Showing results for the entire 300W pulic test set (IBUG dataset, test sets of LFPW and HELEN Average error: 0.0501802344675 AUC @ 0.08: 0.398611514272 Failure rate: 0.088534107402