[Image classification data set] A very comprehensive and practical garbage classification image data set sharing

[Image classification data set] A very comprehensive and practical garbage classification image data set sharing

Dataset introduction:

Training set

The folder structure is as follows (section:

The data under the category 0 folder is displayed as follows (part:

test set

Roughly as follows:

How to get the dataset:

 Summarize:

---

Dataset introduction:

Training set

The garbage classification training set has a total of four categories: recyclables, kitchen waste, hazardous waste, and other wastes.

The garbage classification training set has a total of forty sub-categories, and the labeled data is detailed as follows:

{
    "0": "其他垃圾/一次性快餐盒",
    "1": "其他垃圾/污损塑料",
    "2": "其他垃圾/烟蒂",
    "3": "其他垃圾/牙签",
    "4": "其他垃圾/破碎花盆及碟碗",
    "5": "其他垃圾/竹筷",
    "6": "厨余垃圾/剩饭剩菜",
    "7": "厨余垃圾/大骨头",
    "8": "厨余垃圾/水果果皮",
    "9": "厨余垃圾/水果果肉",
    "10": "厨余垃圾/茶叶渣",
    "11": "厨余垃圾/菜叶菜根",
    "12": "厨余垃圾/蛋壳",
    "13": "厨余垃圾/鱼骨",
    "14": "可回收物/充电宝",
    "15": "可回收物/包",
    "16": "可回收物/化妆品瓶",
    "17": "可回收物/塑料玩具",
    "18": "可回收物/塑料碗盆",
    "19": "可回收物/塑料衣架",
    "20": "可回收物/快递纸袋",
    "21": "可回收物/插头电线",
    "22": "可回收物/旧衣服",
    "23": "可回收物/易拉罐",
    "24": "可回收物/枕头",
    "25": "可回收物/毛绒玩具",
    "26": "可回收物/洗发水瓶",
    "27": "可回收物/玻璃杯",
    "28": "可回收物/皮鞋",
    "29": "可回收物/砧板",
    "30": "可回收物/纸板箱",
    "31": "可回收物/调料瓶",
    "32": "可回收物/酒瓶",
    "33": "可回收物/金属食品罐",
    "34": "可回收物/锅",
    "35": "可回收物/食用油桶",
    "36": "可回收物/饮料瓶",
    "37": "有害垃圾/干电池",
    "38": "有害垃圾/软膏",
    "39": "有害垃圾/过期药物"
}

The folder structure is as follows (section:

The data under the category 0 folder is displayed as follows (part:

test set

In the test set, there are randomly scrambled pictures that have never appeared in the training set. The pictures in it contain 40 categories, which are used for testing after the final model training is completed.

Roughly as follows:

How to get the dataset:

Link: https://pan.baidu.com/s/1hzEJOhd0Q3WjWFIuH-rjhQ 
Extraction code: (Data set is not easy to obtain, the way to obtain it is as follows)

"Like this article if you can, and after bookmarking this article, leave an email in the comment area, and the blogger will send the extraction code immediately."

 Summarize:

        It is not easy to obtain data sets. I heard that the spirit of open source abroad is very popular, so it is not too much for me to open a data set? The overall data set quality is very good, and it is rare to see such a good data set in the whole network.