Image-Cap
说明： 通过机器式学习算法完成对图像的识别，并生成相应的文字序列
(Through the machine learning algorithm, the image recognition is completed and the corresponding text sequence is generated.)
(Through the machine learning algorithm, the image recognition is completed and the corresponding text sequence is generated.)
文件列表:
Image-Cap (0, 2018-03-27)
Image-Cap\.idea (0, 2018-04-11)
Image-Cap\.idea\Image-Cap.iml (455, 2018-03-13)
Image-Cap\.idea\misc.xml (205, 2018-03-13)
Image-Cap\.idea\modules.xml (270, 2018-03-13)
Image-Cap\.idea\workspace.xml (16592, 2018-04-11)
Image-Cap\caption.py (3965, 2018-03-13)
Image-Cap\const.py (129, 2018-03-13)
Image-Cap\data_loader.py (2599, 2018-03-13)
Image-Cap\generate.py (3311, 2018-03-13)
Image-Cap\model.py (6711, 2018-03-13)
Image-Cap\optim.py (1042, 2018-03-13)
Image-Cap\result (0, 2018-03-27)
Image-Cap\result\.DS_Store (6148, 2018-03-13)
Image-Cap\result\000000046378.jpg (133606, 2018-03-13)
Image-Cap\result\000000063740.jpg (109606, 2018-03-13)
Image-Cap\result\000000221754.jpg (154519, 2018-03-13)
Image-Cap\result\000000228214.jpg (170628, 2018-03-13)
Image-Cap\result\000000249025.jpg (126135, 2018-03-13)
Image-Cap\rouge.py (2892, 2018-03-13)
Image-Cap\train.py (10975, 2018-03-13)
Image-Cap\.idea (0, 2018-04-11)
Image-Cap\.idea\Image-Cap.iml (455, 2018-03-13)
Image-Cap\.idea\misc.xml (205, 2018-03-13)
Image-Cap\.idea\modules.xml (270, 2018-03-13)
Image-Cap\.idea\workspace.xml (16592, 2018-04-11)
Image-Cap\caption.py (3965, 2018-03-13)
Image-Cap\const.py (129, 2018-03-13)
Image-Cap\data_loader.py (2599, 2018-03-13)
Image-Cap\generate.py (3311, 2018-03-13)
Image-Cap\model.py (6711, 2018-03-13)
Image-Cap\optim.py (1042, 2018-03-13)
Image-Cap\result (0, 2018-03-27)
Image-Cap\result\.DS_Store (6148, 2018-03-13)
Image-Cap\result\000000046378.jpg (133606, 2018-03-13)
Image-Cap\result\000000063740.jpg (109606, 2018-03-13)
Image-Cap\result\000000221754.jpg (154519, 2018-03-13)
Image-Cap\result\000000228214.jpg (170628, 2018-03-13)
Image-Cap\result\000000249025.jpg (126135, 2018-03-13)
Image-Cap\rouge.py (2892, 2018-03-13)
Image-Cap\train.py (10975, 2018-03-13)
## Introduction
Image Captioning.
Module implemention from "[Actor-Critic Sequence Training for Image Captioning](https://arxiv.org/abs/1706.09601)".
## Requirement * python 3.5 * pytorch 0.3.1 * numpy 1.13.1 * PIL ## Dataset - COCO Download the coco datas from [here](http://cocodataset.org/#download). We use train2017 and val2017 ## Model training * Step 1 - Download dataset, move train2017/ val2017/ captions_train2017.json and captions_val2017.json to {MODEL_HOME}/data/
* Step 2 - Run caption.py ``` $ python3 caption.py ``` * Step 3 - Train model ``` $ python3 train.py ``` ## Result
## Requirement * python 3.5 * pytorch 0.3.1 * numpy 1.13.1 * PIL ## Dataset - COCO Download the coco datas from [here](http://cocodataset.org/#download). We use train2017 and val2017 ## Model training * Step 1 - Download dataset, move train2017/ val2017/ captions_train2017.json and captions_val2017.json to {MODEL_HOME}/data/
* Step 2 - Run caption.py ``` $ python3 caption.py ``` * Step 3 - Train model ``` $ python3 train.py ``` ## Result
My Captioning: Vase with a flower in it sitting on a table.
My Captioning: Cat is laying on a couch with a stuffed animal.
My Captioning: City street filled with lots of tall buildings.
My Captioning: Desk with a computer and a laptop on it.
My Captioning: Man wearing a black jacket and a white shirt.
近期下载者：
相关文件：
收藏者：