j v d dl Z d dlmZ d dlmZmZmZmZmZ d dl Z
ddl m Z m
Z
ddlmZ G d de ZdS ) N)Path)AnyCallableOptionalTupleUnion ) download_and_extract_archiveverify_str_arg)
VisionDatasetc
e Zd ZdZdZdZ ddeeef ded e e
d
e e
d e d df fd
Z d e
fdZde
d eeef fdZd efdZd e fdZddZ xZS )Food101a `The Food-101 Data Set `_.
The Food-101 is a challenging data set of 101 food categories with 101,000 images.
For each class, 250 manually reviewed test images are provided as well as 750 training images.
On purpose, the training images were not cleaned, and thus still contain some amount of noise.
This comes mostly in the form of intense colors and sometimes wrong labels. All images were
rescaled to have a maximum side length of 512 pixels.
Args:
root (str or ``pathlib.Path``): Root directory of the dataset.
split (string, optional): The dataset split, supports ``"train"`` (default) and ``"test"``.
transform (callable, optional): A function/transform that takes in a PIL image and returns a transformed
version. E.g, ``transforms.RandomCrop``.
target_transform (callable, optional): A function/transform that takes in the target and transforms it.
download (bool, optional): If True, downloads the dataset from the internet and
puts it in root directory. If dataset is already downloaded, it is not
downloaded again. Default is False.
z1http://data.vision.ee.ethz.ch/cvl/food-101.tar.gz 85eeb15f3717b99a5da872d97d918f87trainNFrootsplit transformtarget_transformdownloadreturnc
t ||| t |dd _ t j dz _ j dz _ j dz _ |r
st d g _ g _
t j | dz 5 }t j | }d d d n # 1 swxY w Y t% | _ t+ t- j t/ t1 j _ | D ]M\ }} xj j | gt1 | z z
c_ xj
fd | D z
c_
Nd S )
N)r r r )r testzfood-101metaimagesz;Dataset not found. You can use download=True to download itz.jsonc Z g | ]'} j j | d d (S )z.jpg/)_images_folderjoinpathr ).0 im_rel_pathselfs MC:\Projects\Historical\venv\Lib\site-packages\torchvision/datasets/food101.py
z$Food101.__init__..A sK " " "S^,#,.B.B.B.H.H.M.MN" " " )super__init__r _splitr r _base_folder _meta_folderr _download
_check_exists RuntimeError_labels _image_filesopenjsonloadsreadsortedkeysclassesdictziprangelen class_to_idxitems) r! r r r r r fmetadata class_label im_rel_paths __class__s ` r" r& zFood101.__init__# s EUVVV$UG5FGG OOj 8 - 6 " /( : NN !!## ^ \]] ]
$#ooo5
6
6 ,!z!&&((++H , , , , , , , , , , , , , , , hmmoo.. T\5T\9J9J3K3K!L!L M M)1 )9)9 %K LLT .{ ; *NN5))E 1))%00Ee|r$ c d| j S )Nzsplit=)r' rC s r"
extra_reprzFood101.extra_reprT s % %%%r$ c L t d | j | j fD S )Nc 3 f K | ],}| o| V -d S rB )existsis_dir)r folders r" z(Food101._check_exists..X s7 nn66==??6v}}nnnnnnr$ )allr) r rC s r" r+ zFood101._check_existsW s+ nndFWY]YlEmnnnnnnr$ c t | rd S t | j | j | j d S )N)
download_rootmd5)r+ r
_URLr _MD5rC s r" r* zFood101._downloadZ s< F$TYdiTYWWWWWWr$ )r NNF)r N)__name__
__module__ __qualname____doc__r[ r\ r strr r r boolr& intrD r r rN rP r+ r*
__classcell__)r@ s @r" r r sP ( ?D -D
(,/3
CI H%
#8,
D& & & & &
s
uS#X
&C & & & &ot o o o oX X X X X X X Xr$ r )r0 pathlibr typingr r r r r PIL.ImagerH utilsr
r visionr r r$ r" rk s 8 8 8 8 8 8 8 8 8 8 8 8 8 8 ? ? ? ? ? ? ? ? ! ! ! ! ! !RX RX RX RX RXm RX RX RX RX RXr$