U
(dK @ s d dl Z d dlZd dlZd dlmZmZ d dlmZ d dlmZ d dl Z
d dl Z d dl m
Z
ddlmZ dd lmZ dd
lmZ d ZG d d
d
eeZG dd deZG dd deZG dd deZG dd deZG dd deZdd Zdd Z d d Z dS ) N)ABCabstractmethod)globPath)Image ) _read_png_16 )verify_str_arg)
VisionDataset) KittiFlowSintelFlyingThings3D FlyingChairsHD1Kc sN e Zd ZdZd fdd Zdd Zedd Zd d
Zd d Z d
d Z
Z S ) FlowDatasetFNc s$ t j|d || _g | _g | _d S )N)root)super__init__
transforms
_flow_list _image_list)selfr r __class__ F/tmp/pip-unpacked-wheel-vx7f76es/torchvision/datasets/_optical_flow.pyr s zFlowDataset.__init__c C s" t |}|jdkr |d}|S )NZRGB)r openmodeconvert)r file_nameimgr r r _read_img( s
zFlowDataset._read_imgc C s d S Nr r r! r r r
_read_flow. s zFlowDataset._read_flowc C s | | j| d }| | j| d }| jrT| | j| }| jrN|\}}q\d }nd }}| jd k r~| ||||\}}}}| js|d k r||||fS |||fS d S )Nr r
)r# r r r& _has_builtin_flow_maskr )r indeximg1img2flowvalid_flow_maskr r r __getitem__3 s
zFlowDataset.__getitem__c C s
t | jS r$ )lenr )r r r r __len__J s zFlowDataset.__len__c C s t jj| g| S r$ )torchutilsdataZ
ConcatDataset)r vr r r __rmul__M s zFlowDataset.__rmul__)N) __name__
__module__ __qualname__r' r r# r r& r- r/ r4
__classcell__r r r r r s
r c s6 e Zd ZdZd fdd Z fddZd d
Z ZS ) r a `Sintel `_ Dataset for optical flow.
The dataset is expected to have the following structure: ::
root
Sintel
testing
clean
scene_1
scene_2
...
final
scene_1
scene_2
...
training
clean
scene_1
scene_2
...
final
scene_1
scene_2
...
flow
scene_1
scene_2
...
Args:
root (string): Root directory of the Sintel Dataset.
split (string, optional): The dataset split, either "train" (default) or "test"
pass_name (string, optional): The pass to use, either "clean" (default), "final", or "both". See link above for
details on the different passes.
transforms (callable, optional): A function/transform that takes in
``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
``valid_flow_mask`` is expected for consistency with other datasets which
return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
traincleanNc s t j||d t|ddd t|ddd |dkr`__ dataset for optical flow (2015).
The dataset is expected to have the following structure: ::
root
KittiFlow
testing
image_2
training
image_2
flow_occ
Args:
root (string): Root directory of the KittiFlow Dataset.
split (string, optional): The dataset split, either "train" (default) or "test"
transforms (callable, optional): A function/transform that takes in
``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
Tr9 Nc s t j||d t|ddd t|d |d }ttt|d d }ttt|d d }|rj|srtd
t||D ] \}}| j ||gg7 _ q||d krttt|d d | _
d S )
Nr; r< r= r? r
Zingimage_2z*_10.pngz*_11.pngzZCould not find the Kitti flow images. Please make sure the directory structure is correct.r9 flow_occ) r r r r rH r rI FileNotFoundErrorzipr r )r r r< r images1images2r) r* r r r r s zKittiFlow.__init__c s t |S )a Return example at given index.
Args:
index(int): The index of the example to retrieve
Returns:
tuple: A 4-tuple with ``(img1, img2, flow, valid_flow_mask)``
where ``valid_flow_mask`` is a numpy boolean mask of shape (H, W)
indicating which flow values are valid. The flow is a numpy array of
shape (2, H, W) and the images are PIL images. ``flow`` and ``valid_flow_mask`` are None if
``split="test"``.
rN rO r r r r- s
zKittiFlow.__getitem__c C s t |S r$ )_read_16bits_png_with_flow_and_valid_maskr% r r r r& s zKittiFlow._read_flow)r9 N) r5 r6 r7 rS r' r r- r& r8 r r r r r
s
r
c s6 e Zd ZdZd
fdd Z fddZdd Z ZS ) r a
`FlyingChairs `_ Dataset for optical flow.
You will also need to download the FlyingChairs_train_val.txt file from the dataset page.
The dataset is expected to have the following structure: ::
root
FlyingChairs
data
00001_flow.flo
00001_img1.ppm
00001_img2.ppm
...
FlyingChairs_train_val.txt
Args:
root (string): Root directory of the FlyingChairs Dataset.
split (string, optional): The dataset split, either "train" (default) or "val"
transforms (callable, optional): A function/transform that takes in
``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
``valid_flow_mask`` is expected for consistency with other datasets which
return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
r9 Nc
s t j||d t|ddd t|d }ttt|d d }ttt|d d }d }tj || svt
d
t j t|| t j
d }tt|D ]h}|| } |d kr| d
ks|dkr| dkr| j|| g7 _| j|d| |d| d
gg7 _qd S )Nr; r< )r9 valr? r r2 z*.ppmrE zFlyingChairs_train_val.txtzmThe FlyingChairs_train_val.txt file was not found - please download it from the dataset page (see docstring).Zdtyper9 r
r\ r )r r r r rH r rI rF pathexistsrV npZloadtxtZint32rJ r. r r )
r r r< r imagesflowsZsplit_file_nameZ
split_listrL Zsplit_idr r r r s zFlyingChairs.__init__c s t |S )a Return example at given index.
Args:
index(int): The index of the example to retrieve
Returns:
tuple: A 3-tuple with ``(img1, img2, flow)``.
The flow is a numpy array of shape (2, H, W) and the images are PIL images.
``flow`` is None if ``split="val"``.
If a valid flow mask is generated within the ``transforms`` parameter,
a 4-tuple with ``(img1, img2, flow, valid_flow_mask)`` is returned.
rN rO r r r r- s
zFlyingChairs.__getitem__c C s t |S r$ rP r% r r r r& s zFlyingChairs._read_flow)r9 NrR r r r r r s r c s6 e Zd ZdZd fdd Z fdd Zd
d Z ZS )
r a `FlyingThings3D `_ dataset for optical flow.
The dataset is expected to have the following structure: ::
root
FlyingThings3D
frames_cleanpass
TEST
TRAIN
frames_finalpass
TEST
TRAIN
optical_flow
TEST
TRAIN
Args:
root (string): Root directory of the intel FlyingThings3D Dataset.
split (string, optional): The dataset split, either "train" (default) or "test"
pass_name (string, optional): The pass to use, either "clean" (default) or "final" or "both". See link above for
details on the different passes.
camera (string, optional): Which camera to return images from. Can be either "left" (default) or "right" or "both".
transforms (callable, optional): A function/transform that takes in
``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
``valid_flow_mask`` is expected for consistency with other datasets which
return a built-in valid mask, such as :class:`~torchvision.datasets.KittiFlow`.
r9 r: leftNc s t j||d t|ddd | }t|ddd dgdgddgd| }t d d
d d krjd d
gn g}t|d }d}t|||D ]\\} ttt || | d } t fdd| D } ttt |d | d }
t fdd|
D }
| r |
st
dt | |
D ]\} } ttt | d }
ttt | d }t t
|d D ]}dkr| j|
| |
|d gg7 _| j|| g7 _nBdkr^| j|
|d |
| gg7 _| j||d g7 _q^q qd S )Nr; r< r= r? r@ rA Zframes_cleanpassZframes_finalpasscamera)rc rightrC rC rc re r ) into_future into_pastz*/*c 3 s | ]}t | V qd S r$ r ).0 image_dir)rd r r U s z*FlyingThings3D.__init__..Z optical_flowc 3 s | ]}t | V qd S r$ r )rh flow_dirrd directionr r rj X s zcCould not find the FlyingThings3D flow images. Please make sure the directory structure is correct.rD z*.pfmr
rf rg )r r r upperr itertoolsproductrH r rI rV rW rJ r. r r )r r r< r@ rd r rK ZcamerasZ
directionsZ
image_dirsZ flow_dirsri rk ra rb rL r rl r r @ sB
zFlyingThings3D.__init__c s t |S rM rN rO r r r r- k s
zFlyingThings3D.__getitem__c C s t |S r$ ) _read_pfmr% r r r r& z s zFlyingThings3D._read_flow)r9 r: rc NrR r r r r r # s + r c s: e Zd ZdZdZd fdd Zdd Z fd d
Z ZS ) r a `HD1K `__ dataset for optical flow.
The dataset is expected to have the following structure: ::
root
hd1k
hd1k_challenge
image_2
hd1k_flow_gt
flow_occ
hd1k_input
image_2
Args:
root (string): Root directory of the HD1K Dataset.
split (string, optional): The dataset split, either "train" (default) or "test"
transforms (callable, optional): A function/transform that takes in
``img1, img2, flow, valid_flow_mask`` and returns a transformed version.
Tr9 Nc sJ t j||d t|ddd t|d }|dkrtdD ]}ttt|d d |d
d }ttt|d d
|d
d }tt|d D ]8}| j || g7 _ | j
|| ||d gg7 _
qq:nbttt|d d
d }ttt|d d
d } t || D ] \}
} | j
|
| gg7 _
q| j
sFt dd S )Nr; r< r= r? Zhd1kr9 $ Z hd1k_flow_gtrU Z06dz_*.pngZ
hd1k_inputrT r
Zhd1k_challengez*10.pngz*11.pngzTCould not find the HD1K images. Please make sure the directory structure is correct.)
r r r r rJ rH r rI r. r r rW rV ) r r r< r Zseq_idxrb ra rL rX rY Zimage1Zimage2r r r r s$ $$& z
HD1K.__init__c C s t |S r$ rZ r% r r r r& s zHD1K._read_flowc s t |S )a Return example at given index.
Args:
index(int): The index of the example to retrieve
Returns:
tuple: A 4-tuple with ``(img1, img2, flow, valid_flow_mask)`` where ``valid_flow_mask``
is a numpy boolean mask of shape (H, W)
indicating which flow values are valid. The flow is a numpy array of
shape (2, H, W) and the images are PIL images. ``flow`` and ``valid_flow_mask`` are None if
``split="test"``.
rN rO r r r r- s
zHD1K.__getitem__)r9 N) r5 r6 r7 rS r' r r& r- r8 r r r r r ~ s
r c
C s t | d}tj|ddd }|dkr0tdttj|ddd}ttj|ddd}tj|d d
| | d}|||d
d
d dW 5 Q R S Q R X d S )
z#Read .flo file in Middlebury formatrbc )counts PIEHz)Magic number incorrect. Invalid .flo filez.r <>r} r] r r
)ZaxisN)r readlinerstripry rematch Exceptiongroupsfloatr` rw r{ r| ZflipZastyper ) r! r} headerZ dim_matchr r ZscaleZendianr2 r r r rq s$ rq ) ro rF r abcr r r pathlibr r r` r0 ZPILr Zio.imager r1 r Zvisionr __all__r r r
r r r rQ r[ rq r r r r s* 7Q=D[C