fairseq-image-captioning icon indicating copy to clipboard operation
fairseq-image-captioning copied to clipboard

Published 20 hours ago • verified publisher icon krasserm
→

Metadata

Transformer-based image captioning extension for pytorch/fairseq

Results 15 fairseq-image-captioning issues
Sort by recently updated
recently updated
newest added

Train new model using custom data

I want to train a new model using raw images (without bounding boxes) and attached captions to them. Is there any way to do that?

AliOsm avatar AliOsm

Bounding box encoding with sin/cos functions

Inspired by coordinate encoding done in https://arxiv.org/abs/2003.08934.

krasserm avatar krasserm

Implement Meshed-Memory Transformer

https://arxiv.org/abs/1912.08226

krasserm avatar krasserm

Option to fine-tune InceptionV3 encoder

1 comment

krasserm avatar krasserm

Error reporting in processing image caption

![image](https://user-images.githubusercontent.com/44115671/71583036-ab9bac80-2b47-11ea-89e9-6ca0267b183a.png)

SmallSmallQiu avatar SmallSmallQiu

← Metadata

313
Stars
55
Forks
Watchers

Owner

verified publisher icon krasserm

Metadata

Transformer-based image captioning extension for pytorch/fairseq

Back