Dense Image Representation with Spatial Pyramid VLAD Coding of CNN for Locally Robust Captioning

Re-assign community