o
    )iw=                     @   s   d dl mZmZmZmZ d dlZd dlmZmZ d dlm	Z
 d dlmZmZ ddlmZ ddlmZ dd	lmZ G d
d dejZdededededededefddZdee dee deeef fddZG dd dejjZdS )    )DictListOptionalTupleN)nnTensor)
functional)boxesConv2dNormActivation   )_utils)AnchorGenerator)	ImageListc                       sf   e Zd ZdZdZddededdf fdd	Z fd
dZdee	 de
ee	 ee	 f fddZ  ZS )RPNHeada  
    Adds a simple RPN Head with classification and regression heads

    Args:
        in_channels (int): number of channels of the input feature
        num_anchors (int): number of anchors to be predicted
        conv_depth (int, optional): number of convolutions
       r   in_channelsnum_anchorsreturnNc              	      s   t    g }t|D ]}|t||dd d qtj| | _tj||ddd| _	tj||d ddd| _
|  D ] }t|tjrYtjjj|jdd |jd urYtjj|jd q9d S )	N   )kernel_sizeZ
norm_layerr   )r   Zstride   g{Gz?)Zstdr   )super__init__rangeappendr
   r   Z
SequentialconvZConv2d
cls_logits	bbox_predmodules
isinstancetorchinitZnormal_weightbiasZ	constant_)selfr   r   Z
conv_depthZconvs_layer	__class__ o/var/www/html/eduruby.in/lip-sync/lip-sync-env/lib/python3.10/site-packages/torchvision/models/detection/rpn.pyr      s   

zRPNHead.__init__c              	      st   | dd }|d u s|dk r,dD ]}	| d|	 }
| d|	 }|
|v r+||
||< qt ||||||| d S )Nversionr   )r"   r#   zconv.z	conv.0.0.)getpopr   _load_from_state_dict)r$   Z
state_dictprefixZlocal_metadatastrictZmissing_keysZunexpected_keys
error_msgsr+   typeold_keyZnew_keyr'   r)   r*   r.   *   s"   
zRPNHead._load_from_state_dictxc                 C   sD   g }g }|D ]}|  |}|| | || | q||fS N)r   r   r   r   )r$   r4   ZlogitsZbbox_regfeaturetr)   r)   r*   forwardG   s   
zRPNHead.forward)r   )__name__
__module____qualname____doc___versionintr   r.   r   r   r   r8   __classcell__r)   r)   r'   r*   r      s    	.r   r&   NACHWr   c                 C   s6   |  |d|||} | ddddd} | |d|} | S )Nr   r   r   r   r   )viewZpermutereshape)r&   r@   rA   rB   rC   rD   r)   r)   r*   permute_and_flattenQ   s   rH   box_clsbox_regressionc                 C   s   g }g }t | |D ]4\}}|j\}}}}	|jd }
|
d }|| }t||||||	}|| t|||d||	}|| q	tj|dddd} tj|dddd}| |fS )Nr   r   dimr   rE   )zipshaperH   r   r    catflattenrG   )rI   rJ   Zbox_cls_flattenedZbox_regression_flattenedZbox_cls_per_levelZbox_regression_per_levelr@   ZAxCrC   rD   ZAx4rA   rB   r)   r)   r*   concat_box_prediction_layersX   s   

rR   c                       s  e Zd ZdZejejejdZ	d+de	de
jdededed	ed
eeef deeef dededdf fddZdefddZdefddZdee deeeef  deee ee f fddZdedee defddZdededeeeef  dee deee ee f f
d d!Zded"ed#ee d$ee deeef f
d%d&Z	d,d'ed(eeef deeeeef   deee eeef f fd)d*Z  ZS )-RegionProposalNetworkah  
    Implements Region Proposal Network (RPN).

    Args:
        anchor_generator (AnchorGenerator): module that generates the anchors for a set of feature
            maps.
        head (nn.Module): module that computes the objectness and regression deltas
        fg_iou_thresh (float): minimum IoU between the anchor and the GT box so that they can be
            considered as positive during training of the RPN.
        bg_iou_thresh (float): maximum IoU between the anchor and the GT box so that they can be
            considered as negative during training of the RPN.
        batch_size_per_image (int): number of anchors that are sampled during training of the RPN
            for computing the loss
        positive_fraction (float): proportion of positive anchors in a mini-batch during training
            of the RPN
        pre_nms_top_n (Dict[str, int]): number of proposals to keep before applying NMS. It should
            contain two fields: training and testing, to allow for different values depending
            on training or evaluation
        post_nms_top_n (Dict[str, int]): number of proposals to keep after applying NMS. It should
            contain two fields: training and testing, to allow for different values depending
            on training or evaluation
        nms_thresh (float): NMS threshold used for postprocessing the RPN proposals

    )	box_coderproposal_matcherfg_bg_sampler        anchor_generatorheadfg_iou_threshbg_iou_threshbatch_size_per_imagepositive_fractionpre_nms_top_npost_nms_top_n
nms_threshscore_threshr   Nc                    sn   t    || _|| _tjdd| _tj| _	tj
||dd| _t||| _|| _|| _|	| _|
| _d| _d S )N)      ?rb   rb   rb   )weightsT)Zallow_low_quality_matchesgMbP?)r   r   rX   rY   	det_utilsBoxCoderrT   box_opsZbox_ioubox_similarityMatcherrU   BalancedPositiveNegativeSamplerrV   _pre_nms_top_n_post_nms_top_nr`   ra   min_size)r$   rX   rY   rZ   r[   r\   r]   r^   r_   r`   ra   r'   r)   r*   r      s    

zRegionProposalNetwork.__init__c                 C      | j r| jd S | jd S Ntrainingtesting)ro   rj   r$   r)   r)   r*   r^         

z#RegionProposalNetwork.pre_nms_top_nc                 C   rm   rn   )ro   rk   rq   r)   r)   r*   r_      rr   z$RegionProposalNetwork.post_nms_top_nanchorstargetsc                 C   s   g }g }t ||D ]e\}}|d }| dkr2|j}tj|jtj|d}	tj|jd ftj|d}
n2| ||}| |}||j	dd }	|dk}
|
j
tjd}
|| jjk}d|
|< || jjk}d|
|< ||
 ||	 q	||fS )Nr	   r   dtypedevice)min)rv   rW   g      )rN   numelrw   r    ZzerosrO   Zfloat32rg   rU   clamptoZBELOW_LOW_THRESHOLDZBETWEEN_THRESHOLDSr   )r$   rs   rt   labelsmatched_gt_boxesZanchors_per_imageZtargets_per_imageZgt_boxesrw   Zmatched_gt_boxes_per_imageZlabels_per_imageZmatch_quality_matrixZmatched_idxsZ
bg_indicesZinds_to_discardr)   r)   r*   assign_targets_to_anchors   s(   

z/RegionProposalNetwork.assign_targets_to_anchors
objectnessnum_anchors_per_levelc           
      C   sl   g }d}| |dD ]$}|jd }t||  d}|j|dd\}}	||	|  ||7 }q
tj|ddS )Nr   r   rK   )	splitrO   rd   Z	_topk_minr^   Ztopkr   r    rP   )
r$   r   r   roffsetobr   r^   r%   	top_n_idxr)   r)   r*   _get_top_n_idx   s   

z$RegionProposalNetwork._get_top_n_idx	proposalsimage_shapesc                    s  |j d }|j | }||d} fddt|D }t|d}|dd|}| ||}tj	| d}|d d d f }	||	|f }||	|f }||	|f }t
|}
g }g }t||
||D ]]\}}}}t||}t|| j}|| || || }}}t|| jkd }|| || || }}}t|||| j}|d |   }|| || }}|| || qc||fS )Nr   rE   c                    s&   g | ]\}}t j|f|t j d qS )ru   )r    fullZint64).0idxnrw   r)   r*   
<listcomp>   s    z:RegionProposalNetwork.filter_proposals.<locals>.<listcomp>r   r   )rO   rw   detachrG   	enumerater    rP   Z	expand_asr   ZarangeZsigmoidrN   rf   Zclip_boxes_to_imageZremove_small_boxesrl   wherera   Zbatched_nmsr`   r_   r   )r$   r   r   r   r   
num_imagesZlevelsr   Zimage_rangeZ	batch_idxZobjectness_probZfinal_boxesZfinal_scoresr	   scoresZlvlZ	img_shapeZkeepr)   r   r*   filter_proposals   s<   



z&RegionProposalNetwork.filter_proposalspred_bbox_deltasr|   regression_targetsc           
      C   s   |  |\}}ttj|ddd }ttj|ddd }tj||gdd}| }tj|dd}tj|dd}tj|| || ddd|  }t|| || }	|	|fS )a  
        Args:
            objectness (Tensor)
            pred_bbox_deltas (Tensor)
            labels (List[Tensor])
            regression_targets (List[Tensor])

        Returns:
            objectness_loss (Tensor)
            box_loss (Tensor)
        r   rK   gqq?sum)betaZ	reduction)	rV   r    r   rP   rQ   FZsmooth_l1_lossry   Z binary_cross_entropy_with_logits)
r$   r   r   r|   r   Zsampled_pos_indsZsampled_neg_indsZsampled_indsZbox_lossZobjectness_lossr)   r)   r*   compute_loss*  s"   z"RegionProposalNetwork.compute_lossimagesfeaturesc                 C   s   t | }| |\}}| ||}t|}dd |D }dd |D }	t||\}}| j| |}
|
	|dd}
| 
|
||j|	\}}i }| jrr|du rTtd| ||\}}| j||}| ||||\}}||d}||fS )	a=  
        Args:
            images (ImageList): images for which we want to compute the predictions
            features (Dict[str, Tensor]): features computed from the images that are
                used for computing the predictions. Each tensor in the list
                correspond to different feature levels
            targets (List[Dict[str, Tensor]]): ground-truth boxes present in the image (optional).
                If provided, each element in the dict should contain a field `boxes`,
                with the locations of the ground-truth boxes.

        Returns:
            boxes (List[Tensor]): the predicted boxes from the RPN, one Tensor per
                image.
            losses (Dict[str, Tensor]): the losses for the model during training. During
                testing, it is an empty dict.
        c                 S   s   g | ]}|d  j qS )r   )rO   )r   or)   r)   r*   r   l  s    z1RegionProposalNetwork.forward.<locals>.<listcomp>c                 S   s$   g | ]}|d  |d  |d  qS )r   r   r   r)   )r   sr)   r)   r*   r   m  s   $ rE   r   Nztargets should not be None)loss_objectnessloss_rpn_box_reg)listvaluesrY   rX   lenrR   rT   decoder   rF   r   Zimage_sizesro   
ValueErrorr~   encoder   )r$   r   r   rt   r   r   rs   r   Z#num_anchors_per_level_shape_tensorsr   r   r	   r   Zlossesr|   r}   r   r   r   r)   r)   r*   r8   O  s.   zRegionProposalNetwork.forward)rW   r5   )r9   r:   r;   r<   rd   re   rh   ri   __annotations__r   r   Modulefloatr>   r   strr   r^   r_   r   r   r   r~   r   r   r   r   r   r8   r?   r)   r)   r'   r*   rS   q   s    


%
&
9

)
rS   )typingr   r   r   r   r    r   r   Ztorch.nnr   r   Ztorchvision.opsr	   rf   r
    r   rd   Zanchor_utilsr   Z
image_listr   r   r   r>   rH   rR   rS   r)   r)   r)   r*   <module>   s    &B&