Dynamic Memory Networks for Visual and Textual Question Answering

Neural network architectures with memory and attention mechanisms exhibit certain reasoning capabilities required for question answering. One such architecture, the dynamic memory network (DMN), obtained high accuracy on a variety of language tasks... (read more)

PDF Abstract
TASK DATASET MODEL METRIC NAME METRIC VALUE GLOBAL RANK RESULT BENCHMARK
Visual Question Answering COCO Visual Question Answering (VQA) real images 1.0 open ended DMN+ [xiong2016dynamic] Percentage correct 60.4 # 8
Visual Question Answering VQA v1 test-dev DMN+ Accuracy 60.3 # 6
Visual Question Answering VQA v1 test-std DMN+ Accuracy 60.4 # 4

Methods used in the Paper


METHOD TYPE
Softmax
Output Functions
GRU
Recurrent Neural Networks
Dynamic Memory Network
Working Memory Models
Memory Network
Working Memory Models