Open Access

Table 6

The comparison of object goal navigation models

Model Observation Target Scene graph Memory Learning method
TD-A3C[2] ResNet-50 Glove None LSTM A3C
Bayes[68] ResNet-50 Glove None LSTM A3C
SAVN[69] ResNet-18 Glove None ML A3C+MAML
SpAtt[70] ResNet-18 Glove None ML A3C+MAML
SP[25] ResNet-50 Glove VG+GCN LSTM A3C
GVE[57] ResNet-18 Glove VG+GTN ML A3C+MAML
VGM[39] CNN Glove VGM+GCN IL+LSTM A3C
ORG[27] ResNet-18+Faster-RCNN Glove ORG+GCN LSTM A3C+IL
HOZ[34] ResNet-18+Faster-RCNN Glove HOZ+GCN LSTM A3C+MAML
DOA[31] ResNet-18+Faster-RCNN Glove DOA+GCN LSTM A3C
ABM[49] ResNet-18+DETR Glove ARG+GAT ML A3C
VTNet[71] ResNet-18+DETR Glove VTN LSTM A3C
GTV+MTL[73] ResNet-18 Glove CG+GTV ML A3C+MTL

Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.

Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.

Initial download of the metrics may take a while.