AI Paper Atlas

Research map / Vision-language and multimodal / Visual question answering

Visual Question Generation for Class Acquisition of Unknown Objects

Kohei Uehara, Antonio Tejero-De-Pablos, Yoshitaka Ushiku, Tatsuya Harada

ECCV 2018 · Vision-language and multimodal › Visual question answering · approach: Attention and prior-robust VQA · 18 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use softmax entropy instead of least-confident threshold for unknown detectionResNet152 classifier · Unknown object classification (ILSVRC2012 known vs ILSVRC2010 unknown)F measure: 0.635 -> 0.654Improved
use softmax entropy instead of least-confident threshold for unknown detectionCaffeNet classifier · Unknown object classification (ILSVRC2012 known vs ILSVRC2010 unknown)F measure: 0.522 -> 0.526No clear effect
use softmax entropy instead of Bendale et al. for unknown detectionVGGNet classifier · Unknown object classification (ILSVRC2012 known vs ILSVRC2010 unknown)F measure: 0.553 -> 0.602Improved
use a stronger classifier (CaffeNet to VGGNet to ResNet)entropy-based unknown detection · Unknown object classification (ILSVRC2012 known vs ILSVRC2010 unknown)F measure: 0.526 -> 0.602 -> 0.654Improved
use entropy rather than Bendale et al. for speedCaffeNet classifier · Unknown classification execution timetime (sec/image): 15.6 -> 0.0400Improved
use entropy rather than least confidentCaffeNet classifier · Unknown classification execution timetime (sec/image): 0.0365 -> 0.0400No clear effect
add target region features and hypernym target word to the decoder (vs whole-image CNN + LSTM)LSTM question generator · Visual Genome question generation (test)BLEU-4: 0.110 -> 0.175Improved
add target region features and hypernym target word to the decoder (vs whole-image CNN + LSTM)LSTM question generator · Visual Genome question generation (test)METEOR: 0.163 -> 0.197Improved
generate questions with the proposed decoder instead of retrievalquestion generation · Visual Genome question generation (test)BLEU-1: 0.438 -> 0.518Improved
add target conditioning to the LSTM decoderquestion generation · Human evaluation, naturalness (AMT)naturalness scoreNo clear effect
add target conditioning to the LSTM decoderquestion generation · Human evaluation, relevance to target region (AMT)relevance scoreImproved
use top k=3 instead of k=2 classes for the common hypernym target word (ablation)unknown object VQG pipeline · Human evaluation, relevance (AMT)average relevance scoreMixed

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy