Transformers documentation

DPT

Join the Hugging Face community

to get started

# DPT

## Overview

The DPT model was proposed in Vision Transformers for Dense Prediction by René Ranftl, Alexey Bochkovskiy, Vladlen Koltun. DPT is a model that leverages the Vision Transformer (ViT) as backbone for dense prediction tasks like semantic segmentation and depth estimation.

The abstract from the paper is the following:

We introduce dense vision transformers, an architecture that leverages vision transformers in place of convolutional networks as a backbone for dense prediction tasks. We assemble tokens from various stages of the vision transformer into image-like representations at various resolutions and progressively combine them into full-resolution predictions using a convolutional decoder. The transformer backbone processes representations at a constant and relatively high resolution and has a global receptive field at every stage. These properties allow the dense vision transformer to provide finer-grained and more globally coherent predictions when compared to fully-convolutional networks. Our experiments show that this architecture yields substantial improvements on dense prediction tasks, especially when a large amount of training data is available. For monocular depth estimation, we observe an improvement of up to 28% in relative performance when compared to a state-of-the-art fully-convolutional network. When applied to semantic segmentation, dense vision transformers set a new state of the art on ADE20K with 49.02% mIoU. We further show that the architecture can be fine-tuned on smaller datasets such as NYUv2, KITTI, and Pascal Context where it also sets the new state of the art.

DPT architecture. Taken from the original paper.

This model was contributed by nielsr. The original code can be found here.

## DPTConfig

### class transformers.DPTConfig

< >

( hidden_size = 768 num_hidden_layers = 12 num_attention_heads = 12 intermediate_size = 3072 hidden_act = 'gelu' hidden_dropout_prob = 0.0 attention_probs_dropout_prob = 0.0 initializer_range = 0.02 layer_norm_eps = 1e-12 image_size = 384 patch_size = 16 num_channels = 3 qkv_bias = True backbone_out_indices = [2, 5, 8, 11] readout_type = 'project' reassemble_factors = [4, 2, 1, 0.5] neck_hidden_sizes = [96, 192, 384, 768] fusion_hidden_size = 256 head_in_index = -1 use_batch_norm_in_fusion_residual = False use_auxiliary_head = True auxiliary_loss_weight = 0.4 semantic_loss_ignore_index = 255 semantic_classifier_dropout = 0.1 **kwargs )

Parameters

• hidden_size (int, optional, defaults to 768) — Dimensionality of the encoder layers and the pooler layer.
• num_hidden_layers (int, optional, defaults to 12) — Number of hidden layers in the Transformer encoder.
• num_attention_heads (int, optional, defaults to 12) — Number of attention heads for each attention layer in the Transformer encoder.
• intermediate_size (int, optional, defaults to 3072) — Dimensionality of the “intermediate” (i.e., feed-forward) layer in the Transformer encoder.
• hidden_act (str or function, optional, defaults to "gelu") — The non-linear activation function (function or string) in the encoder and pooler. If string, "gelu", "relu", "selu" and "gelu_new" are supported.
• hidden_dropout_prob (float, optional, defaults to 0.1) — The dropout probabilitiy for all fully connected layers in the embeddings, encoder, and pooler.
• attention_probs_dropout_prob (float, optional, defaults to 0.1) — The dropout ratio for the attention probabilities.
• initializer_range (float, optional, defaults to 0.02) — The standard deviation of the truncated_normal_initializer for initializing all weight matrices.
• layer_norm_eps (float, optional, defaults to 1e-12) — The epsilon used by the layer normalization layers.
• image_size (int, optional, defaults to 384) — The size (resolution) of each image.
• patch_size (int, optional, defaults to 16) — The size (resolution) of each patch.
• num_channels (int, optional, defaults to 3) — The number of input channels.
• qkv_bias (bool, optional, defaults to True) — Whether to add a bias to the queries, keys and values.
• backbone_out_indices (List[int], optional, defaults to [2, 5, 8, 11]) — Indices of the intermediate hidden states to use from backbone.
• readout_type (str, optional, defaults to "project") — The readout type to use when processing the readout token (CLS token) of the intermediate hidden states of the ViT backbone. Can be one of ["ignore", "add", "project"].

• “ignore” simply ignores the CLS token.
• “add” passes the information from the CLS token to all other tokens by adding the representations.
• “project” passes information to the other tokens by concatenating the readout to all other tokens before projecting the representation to the original feature dimension D using a linear layer followed by a GELU non-linearity.
• reassemble_factors (List[int], optional, defaults to [4, 2, 1, 0.5]) — The up/downsampling factors of the reassemble layers.
• neck_hidden_sizes (List[str], optional, defaults to [96, 192, 384, 768]) — The hidden sizes to project to for the feature maps of the backbone.
• fusion_hidden_size (int, optional, defaults to 256) — The number of channels before fusion.
• head_in_index (int, optional, defaults to -1) — The index of the features to use in the heads.
• use_batch_norm_in_fusion_residual (bool, optional, defaults to False) — Whether to use batch normalization in the pre-activate residual units of the fusion blocks.
• use_auxiliary_head (bool, optional, defaults to True) — Whether to use an auxiliary head during training.
• auxiliary_loss_weight (float, optional, defaults to 0.4) — Weight of the cross-entropy loss of the auxiliary head.
• semantic_loss_ignore_index (int, optional, defaults to 255) — The index that is ignored by the loss function of the semantic segmentation model.
• semantic_classifier_dropout (float, optional, defaults to 0.1) — The dropout ratio for the semantic classification head.

This is the configuration class to store the configuration of a DPTModel. It is used to instantiate an DPT model according to the specified arguments, defining the model architecture. Instantiating a configuration with the defaults will yield a similar configuration to that of the DPT Intel/dpt-large architecture.

Configuration objects inherit from PretrainedConfig and can be used to control the model outputs. Read the documentation from PretrainedConfig for more information.

Example:

>>> from transformers import DPTModel, DPTConfig

>>> # Initializing a DPT dpt-large style configuration
>>> configuration = DPTConfig()

>>> # Initializing a model from the dpt-large style configuration
>>> model = DPTModel(configuration)

>>> # Accessing the model configuration
>>> configuration = model.config

## DPTFeatureExtractor

### class transformers.DPTFeatureExtractor

< >

( do_resize = True size = 384 keep_aspect_ratio = False ensure_multiple_of = 1 resample = <Resampling.BILINEAR: 2> do_normalize = True image_mean = None image_std = None **kwargs )

Parameters

• do_resize (bool, optional, defaults to True) — Whether to resize the input to a certain size.
• size (‘int’ or Tuple(int), optional, defaults to 384) — Resize the input to the given size. If a tuple is provided, it should be (width, height). If only an integer is provided, then the input will be resized to (size, size). Only has an effect if do_resize is set to True.
• ensure_multiple_of (int, optional, defaults to 1) — Ensure that the input is resized to a multiple of this value. Only has an effect if do_resize is set to True.
• keep_aspect_ratio (bool, optional, defaults to False) — Whether to keep the aspect ratio of the input. Only has an effect if do_resize is set to True.
• resample (int, optional, defaults to PIL.Image.BILINEAR) — An optional resampling filter. This can be one of PIL.Image.NEAREST, PIL.Image.BOX, PIL.Image.BILINEAR, PIL.Image.HAMMING, PIL.Image.BICUBIC or PIL.Image.LANCZOS. Only has an effect if do_resize is set to True.
• do_normalize (bool, optional, defaults to True) — Whether or not to normalize the input with mean and standard deviation.
• image_mean (List[int], defaults to [0.5, 0.5, 0.5]) — The sequence of means for each channel, to be used when normalizing images.
• image_std (List[int], defaults to [0.5, 0.5, 0.5]) — The sequence of standard deviations for each channel, to be used when normalizing images.

Constructs a DPT feature extractor.

This feature extractor inherits from FeatureExtractionMixin which contains most of the main methods. Users should refer to this superclass for more information regarding those methods.

#### __call__

< >

( images: typing.Union[PIL.Image.Image, numpy.ndarray, ForwardRef('torch.Tensor'), typing.List[PIL.Image.Image], typing.List[numpy.ndarray], typing.List[ForwardRef('torch.Tensor')]] return_tensors: typing.Union[str, transformers.utils.generic.TensorType, NoneType] = None **kwargs ) BatchFeature

Parameters

• images (PIL.Image.Image, np.ndarray, torch.Tensor, List[PIL.Image.Image], List[np.ndarray], List[torch.Tensor]) — The image or batch of images to be prepared. Each image can be a PIL image, NumPy array or PyTorch tensor. In case of a NumPy array/PyTorch tensor, each image should be of shape (C, H, W), where C is a number of channels, H and W are image height and width.
• return_tensors (str or TensorType, optional, defaults to 'np') — If set, will return tensors of a particular framework. Acceptable values are:

• 'tf': Return TensorFlow tf.constant objects.
• 'pt': Return PyTorch torch.Tensor objects.
• 'np': Return NumPy np.ndarray objects.
• 'jax': Return JAX jnp.ndarray objects.

Returns

BatchFeature

A BatchFeature with the following fields:

• pixel_values — Pixel values to be fed to a model, of shape (batch_size, num_channels, height, width).

Main method to prepare for the model one or several image(s).

NumPy arrays and PyTorch tensors are converted to PIL images when resizing, so the most efficient is to pass PIL images.

## DPTModel

### class transformers.DPTModel

< >

( config add_pooling_layer = True )

Parameters

• config (ViTConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

The bare DPT Model transformer outputting raw hidden-states without any specific head on top. This model is a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

#### forward

< >

( pixel_values head_mask = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.BaseModelOutputWithPooling or tuple(torch.FloatTensor)

Parameters

• pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using ViTFeatureExtractor. See ViTFeatureExtractor.call() for details.
• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.
• output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.
• return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.

Returns

transformers.modeling_outputs.BaseModelOutputWithPooling or tuple(torch.FloatTensor)

A transformers.modeling_outputs.BaseModelOutputWithPooling or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (DPTConfig) and inputs.

• last_hidden_state (torch.FloatTensor of shape (batch_size, sequence_length, hidden_size)) — Sequence of hidden-states at the output of the last layer of the model.

• pooler_output (torch.FloatTensor of shape (batch_size, hidden_size)) — Last layer hidden-state of the first token of the sequence (classification token) after further processing through the layers used for the auxiliary pretraining task. E.g. for BERT-family of models, this returns the classification token after processing through a linear layer and a tanh activation function. The linear layer weights are trained from the next sentence prediction (classification) objective during pretraining.

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, sequence_length, hidden_size).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, sequence_length, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

The DPTModel forward method, overrides the __call__ special method.

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Example:

>>> from transformers import DPTFeatureExtractor, DPTModel
>>> import torch

>>> image = dataset["test"]["image"][0]

>>> feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-large")
>>> model = DPTModel.from_pretrained("Intel/dpt-large")

>>> inputs = feature_extractor(image, return_tensors="pt")

...     outputs = model(**inputs)

>>> last_hidden_states = outputs.last_hidden_state
>>> list(last_hidden_states.shape)
[1, 577, 1024]

## DPTForDepthEstimation

### class transformers.DPTForDepthEstimation

< >

( config )

Parameters

• config (ViTConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

DPT Model with a depth estimation head on top (consisting of 3 convolutional layers) e.g. for KITTI, NYUv2.

This model is a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

#### forward

< >

( pixel_values head_mask = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.DepthEstimatorOutput or tuple(torch.FloatTensor)

Parameters

• pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using ViTFeatureExtractor. See ViTFeatureExtractor.call() for details.
• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.
• output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.
• return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
• labels (torch.LongTensor of shape (batch_size, height, width), optional) — Ground truth depth estimation maps for computing the loss.

Returns

transformers.modeling_outputs.DepthEstimatorOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.DepthEstimatorOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (DPTConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) — Classification (or regression if config.num_labels==1) loss.

• predicted_depth (torch.FloatTensor of shape (batch_size, height, width)) — Predicted depth for each pixel.

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, num_channels, height, width).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, patch_size, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

The DPTForDepthEstimation forward method, overrides the __call__ special method.

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Examples:

>>> from transformers import DPTFeatureExtractor, DPTForDepthEstimation
>>> import torch
>>> import numpy as np
>>> from PIL import Image
>>> import requests

>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
>>> image = Image.open(requests.get(url, stream=True).raw)

>>> feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-large")
>>> model = DPTForDepthEstimation.from_pretrained("Intel/dpt-large")

>>> # prepare image for the model
>>> inputs = feature_extractor(images=image, return_tensors="pt")

...     outputs = model(**inputs)
...     predicted_depth = outputs.predicted_depth

>>> # interpolate to original size
>>> prediction = torch.nn.functional.interpolate(
...     predicted_depth.unsqueeze(1),
...     size=image.size[::-1],
...     mode="bicubic",
...     align_corners=False,
... )

>>> # visualize the prediction
>>> output = prediction.squeeze().cpu().numpy()
>>> formatted = (output * 255 / np.max(output)).astype("uint8")
>>> depth = Image.fromarray(formatted)

## DPTForSemanticSegmentation

### class transformers.DPTForSemanticSegmentation

< >

( config )

Parameters

• config (ViTConfig) — Model configuration class with all the parameters of the model. Initializing with a config file does not load the weights associated with the model, only the configuration. Check out the from_pretrained() method to load the model weights.

DPT Model with a semantic segmentation head on top e.g. for ADE20k, CityScapes.

This model is a PyTorch torch.nn.Module subclass. Use it as a regular PyTorch Module and refer to the PyTorch documentation for all matter related to general usage and behavior.

#### forward

< >

( pixel_values = None head_mask = None labels = None output_attentions = None output_hidden_states = None return_dict = None ) transformers.modeling_outputs.SemanticSegmenterOutput or tuple(torch.FloatTensor)

Parameters

• pixel_values (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — Pixel values. Pixel values can be obtained using ViTFeatureExtractor. See ViTFeatureExtractor.call() for details.
• head_mask (torch.FloatTensor of shape (num_heads,) or (num_layers, num_heads), optional) — Mask to nullify selected heads of the self-attention modules. Mask values selected in [0, 1]:

• output_attentions (bool, optional) — Whether or not to return the attentions tensors of all attention layers. See attentions under returned tensors for more detail.
• output_hidden_states (bool, optional) — Whether or not to return the hidden states of all layers. See hidden_states under returned tensors for more detail.
• return_dict (bool, optional) — Whether or not to return a ModelOutput instead of a plain tuple.
• labels (torch.LongTensor of shape (batch_size, height, width), optional) — Ground truth semantic segmentation maps for computing the loss. Indices should be in [0, ..., config.num_labels - 1]. If config.num_labels > 1, a classification loss is computed (Cross-Entropy).

Returns

transformers.modeling_outputs.SemanticSegmenterOutput or tuple(torch.FloatTensor)

A transformers.modeling_outputs.SemanticSegmenterOutput or a tuple of torch.FloatTensor (if return_dict=False is passed or when config.return_dict=False) comprising various elements depending on the configuration (DPTConfig) and inputs.

• loss (torch.FloatTensor of shape (1,), optional, returned when labels is provided) — Classification (or regression if config.num_labels==1) loss.

• logits (torch.FloatTensor of shape (batch_size, config.num_labels, logits_height, logits_width)) — Classification scores for each pixel.

The logits returned do not necessarily have the same size as the pixel_values passed as inputs. This is to avoid doing two interpolations and lose some quality when a user needs to resize the logits to the original image size as post-processing. You should always check your logits shape and resize as needed.

• hidden_states (tuple(torch.FloatTensor), optional, returned when output_hidden_states=True is passed or when config.output_hidden_states=True) — Tuple of torch.FloatTensor (one for the output of the embeddings, if the model has an embedding layer, + one for the output of each layer) of shape (batch_size, patch_size, hidden_size).

Hidden-states of the model at the output of each layer plus the optional initial embedding outputs.

• attentions (tuple(torch.FloatTensor), optional, returned when output_attentions=True is passed or when config.output_attentions=True) — Tuple of torch.FloatTensor (one for each layer) of shape (batch_size, num_heads, patch_size, sequence_length).

Attentions weights after the attention softmax, used to compute the weighted average in the self-attention heads.

The DPTForSemanticSegmentation forward method, overrides the __call__ special method.

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the pre and post processing steps while the latter silently ignores them.

Examples:

>>> from transformers import DPTFeatureExtractor, DPTForSemanticSegmentation
>>> from PIL import Image
>>> import requests

>>> url = "http://images.cocodataset.org/val2017/000000039769.jpg"
>>> image = Image.open(requests.get(url, stream=True).raw)

>>> logits = outputs.logits