Beyond a Pre-Trained Object Detector: Cross-Modal Textual and Visual Context for Image Captioning

Benchmark Model Rank Results
image-captioning-on-coco-captionsXmodal-Ctx#11BLEU-4: 41.4CIDER: 139.9METEOR: 30.4SPICE: 24.0ROUGE-L: 60.4…
image-captioning-on-coco-captionsXmodal-Ctx + OSCAR#13BLEU-4: 41.3CIDER: 142.2SPICE: 24.9
image-captioning-on-coco-captionsXmodal-Ctx#21BLEU-4: 39.7CIDER: 135.9METEOR: 30.0SPICE: 23.7ROUGE-L: 59.5…