vsubhashini / noc Goto Github PK

View Code? Open in Web Editor NEW

41.0 41.0 23.0 37.73 MB

Novel Object Captioner - Captioning Images with diverse objects

License: Other

CMake 2.98% Makefile 0.67% Shell 0.41% Python 8.55% C++ 81.61% Cuda 4.81% MATLAB 0.97%

noc's People

Contributors

Stargazers

Watchers

noc's Issues

I want to know how to train the model

hi
I want to reproduce your experiment，in the code to generate .h5 file
POOLFEAT_FILE_PATTERN = 'data/coco2014/coco2014_{0}vgg_fc7.txt'
SENTS_FILE_PATTERN = 'data/coco2014/sents/coco_sentences{0}tokens.txt'
LABEL_FILE_PATTERN = 'data/coco2014/sents/labels_glove72k{0}.txt' #train2014

IMAGEID_FILE_PATTERN = 'data/coco2014/coco_rm8objs_image_list_{0}.txt'

IMAGEID_FILE_PATTERN = 'data/coco2014/cvpr17_rm8newobjs/coco_rm8newobjs_image_list_{0}.txt'
i want to know how can i get these handled documents like coco2014_{0}_vgg_fc7.txt?
I am a beginner，I am so sorry to take up your time. Can you tell me how to reproduce your experiment?
thank you!

AssertionError when running noc_captioner.py -i images_list.txt

Computing features for images 0-9 of 10
Traceback (most recent call last):
  File "noc_captioner.py", line 569, in <module>
    main()
  File "noc_captioner.py", line 553, in main
    strategies=STRATEGIES, display_vocab=vocab_list)
  File "noc_captioner.py", line 286, in run_pred_iters
    run_pred_iter(pred_net, image_fc7, display_vocab, strategies=strategies)
  File "noc_captioner.py", line 199, in run_pred_iter
    captions, probs = predict_image_caption(net, mean_pool_fc7, vocab_list, strategy=strategy)
  File "noc_captioner.py", line 97, in predict_image_caption
    return predict_image_caption_beam_search(net, mean_pool_fc7, vocab_list, strategy)
  File "noc_captioner.py", line 174, in predict_image_caption_beam_search
    assert probs.shape[0] == len(vocab_list)
AssertionError

I'm wondering how you evaluate the F1-score of your model

I find "caption-eval" on your homepage, but I still have no idea of evaluate the F1-score. I'm appreciated it if you can provide the code or something can help.

How can I define the global name 'FeatureExtractor'?

I'm sorry to bother you, but I'm really confused about this.
As I'm trying to run the noc_captioner.py, there is an error:

Setting up CNN... Traceback (most recent call last): File "noc_captioner.py", line 571, in <module> main() File "noc_captioner.py", line 511, in main feature_extractor = FeatureExtractor(args.vggmodel, VGG_NET_FILE, DEVICE_ID) NameError: global name 'FeatureExtractor' is not defined
Am I lost something necessary for the captioner? How can I fix it? I really appreciate if you can help me.

vsubhashini / noc Goto Github PK

noc's People

Contributors

Stargazers

Watchers

Forkers

noc's Issues

I want to know how to train the model

IMAGEID_FILE_PATTERN = 'data/coco2014/coco_rm8objs_image_list_{0}.txt'

AssertionError when running noc_captioner.py -i images_list.txt

I'm wondering how you evaluate the F1-score of your model

How can I define the global name 'FeatureExtractor'?

Recommend Projects

React

Vue.js

Typescript

TensorFlow

Django

Laravel

D3

Recommend Topics

javascript

web

server

Machine learning

Visualization

Game

Recommend Org

Facebook

Microsoft

Google

Alibaba

D3

Tencent