Transformers documentation

MobileViT

Hugging Face's logo
Join the Hugging Face community

and get access to the augmented documentation experience

to get started

MobileViT

Overview

The MobileViT model was proposed in MobileViT: Light-weight, General-purpose, and Mobile-friendly Vision Transformer by Sachin Mehta and Mohammad Rastegari. MobileViT introduces a new layer that replaces local processing in convolutions with global processing using transformers.

The abstract from the paper is the following:

Light-weight convolutional neural networks (CNNs) are the de-facto for mobile vision tasks. Their spatial inductive biases allow them to learn representations with fewer parameters across different vision tasks. However, these networks are spatially local. To learn global representations, self-attention-based vision trans-formers (ViTs) have been adopted. Unlike CNNs, ViTs are heavy-weight. In this paper, we ask the following question: is it possible to combine the strengths of CNNs and ViTs to build a light-weight and low latency network for mobile vision tasks? Towards this end, we introduce MobileViT, a light-weight and general-purpose vision transformer for mobile devices. MobileViT presents a different perspective for the global processing of information with transformers, i.e., transformers as convolutions. Our results show that MobileViT significantly outperforms CNN- and ViT-based networks across different tasks and datasets. On the ImageNet-1k dataset, MobileViT achieves top-1 accuracy of 78.4% with about 6 million parameters, which is 3.2% and 6.2% more accurate than MobileNetv3 (CNN-based) and DeIT (ViT-based) for a similar number of parameters. On the MS-COCO object detection task, MobileViT is 5.7% more accurate than MobileNetv3 for a similar number of parameters.

Tips:

  • MobileViT is more like a CNN than a Transformer model. It does not work on sequence data but on batches of images. Unlike ViT, there are no embeddings. The backbone model outputs a feature map.
  • One can use MobileViTFeatureExtractor to prepare images for the model. Note that if you do your own preprocessing, the pretrained checkpoints expect images to be in BGR pixel order (not RGB).
  • The available image classification checkpoints are pre-trained on ImageNet-1k (also referred to as ILSVRC 2012, a collection of 1.3 million images and 1,000 classes).
  • The segmentation model uses a DeepLabV3 head. The available semantic segmentation checkpoints are pre-trained on PASCAL VOC.

This model was contributed by matthijs. The original code and weights can be found here.

MobileViTConfig

class transformers.MobileViTConfig

< >

( num_channels = 3 image_size = 256 patch_size = 2 hidden_sizes = [144, 192, 240] neck_hidden_sizes = [16, 32, 64, 96, 128, 160, 640] num_attention_heads = 4 mlp_ratio = 2.0 expand_ratio = 4.0 hidden_act = 'silu' conv_kernel_size = 3 output_stride = 32 hidden_dropout_prob = 0.1 attention_probs_dropout_prob = 0.0 classifier_dropout_prob = 0.1 initializer_range = 0.02 layer_norm_eps = 1e-05 qkv_bias = True aspp_out_channels = 256 atrous_rates = [6, 12, 18] aspp_dropout_prob = 0.1 semantic_loss_ignore_index = 255 **kwargs )

Parameters

  • num_channels (int, optional, defaults to 3) — The number of input channels.
  • image_size (int, optional, defaults to 256) — The size (resolution) of each image.
  • patch_size (int, optional, defaults to 2) — The size (resolution) of each patch.
  • hidden_sizes (List[int], optional, defaults to [144, 192, 240]) — Dimensionality (hidden size) of the Transformer encoders at each stage.
  • neck_hidden_sizes (List[int], optional, defaults to [16, 32, 64, 96, 128, 160, 640]) — The number of channels for the feature maps of the backbone.
  • num_attention_heads (int, optional, defaults to 4) — Number of attention heads for each attention layer in the Transformer encoder.
  • mlp_ratio (float, optional, defaults to 2.0) — The ratio of the number of channels in the output of the MLP to the number of channels in the input.
  • expand_ratio (float, optional, defaults to 4.0) — Expansion factor for the MobileNetv2 layers.
  • hidden_act (str or function, optional, defaults to "silu") — The non-linear activation function (function or string) in the Transformer encoder and convolution layers.
  • conv_kernel_size (int, optional, defaults to 3) — The size of the convolutional kernel in the MobileViT layer.
  • output_stride (int, optional, defaults to 32) — The ratio of the spatial resolution of the output to the resolution of the input image.
  • hidden_dropout_prob (float, optional, defaults to 0.1) — The dropout probabilitiy for all fully connected layers in the Transformer encoder.
  • attention_probs_dropout_prob (float, optional, defaults to 0.0) — The dropout ratio for the attention probabilities.
  • classifier_dropout_prob (float, optional, defaults to 0.1) — The dropout ratio for attached classifiers.
  • 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-5) — The epsilon used by the layer normalization layers.
  • qkv_bias (bool, optional, defaults to True) — Whether to add a bias to the queries, keys and values.
  • aspp_out_channels (int, optional, defaults to 256) — Number of output channels used in the ASPP layer for semantic segmentation.
  • atrous_rates (List[int], optional, defaults to [6, 12, 18]) — Dilation (atrous) factors used in the ASPP layer for semantic segmentation.
  • aspp_dropout_prob (float, optional, defaults to 0.1) — The dropout ratio for the ASPP layer for semantic segmentation.
  • semantic_loss_ignore_index (int, optional, defaults to 255) — The index that is ignored by the loss function of the semantic segmentation model.

This is the configuration class to store the configuration of a MobileViTModel. It is used to instantiate a MobileViT 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 MobileViT apple/mobilevit-small 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 MobileViTConfig, MobileViTModel

>>> # Initializing a mobilevit-small style configuration
>>> configuration = MobileViTConfig()

>>> # Initializing a model from the mobilevit-small style configuration
>>> model = MobileViTModel(configuration)

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

MobileViTFeatureExtractor

class transformers.MobileViTFeatureExtractor

< >

( do_resize = True size = 288 resample = <Resampling.BILINEAR: 2> do_center_crop = True crop_size = 256 do_flip_channel_order = True **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 288) — 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 match the shorter side. 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_center_crop (bool, optional, defaults to True) — Whether to crop the input at the center. If the input size is smaller than crop_size along any edge, the image is padded with 0’s and then center cropped.
  • crop_size (int, optional, defaults to 256) — Desired output size when applying center-cropping. Only has an effect if do_center_crop is set to True.
  • do_flip_channel_order (bool, optional, defaults to True) — Whether to flip the color channels from RGB to BGR.

Constructs a MobileViT 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.

MobileViTModel

class transformers.MobileViTModel

< >

( config: MobileViTConfig expand_output: bool = True )

Parameters

  • config (MobileViTConfig) — 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 MobileViT model 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: typing.Optional[torch.Tensor] = None output_hidden_states: typing.Optional[bool] = None return_dict: typing.Optional[bool] = None ) → transformers.modeling_outputs.BaseModelOutputWithPoolingAndNoAttention 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 MobileViTFeatureExtractor. See MobileViTFeatureExtractor.call() for details.
  • 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.BaseModelOutputWithPoolingAndNoAttention or tuple(torch.FloatTensor)

A transformers.modeling_outputs.BaseModelOutputWithPoolingAndNoAttention 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 (MobileViTConfig) and inputs.

  • last_hidden_state (torch.FloatTensor of shape (batch_size, num_channels, height, width)) — 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 after a pooling operation on the spatial dimensions.

  • 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.

The MobileViTModel 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 MobileViTFeatureExtractor, MobileViTModel
>>> import torch
>>> from datasets import load_dataset

>>> dataset = load_dataset("huggingface/cats-image")
>>> image = dataset["test"]["image"][0]

>>> feature_extractor = MobileViTFeatureExtractor.from_pretrained("apple/mobilevit-small")
>>> model = MobileViTModel.from_pretrained("apple/mobilevit-small")

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

>>> with torch.no_grad():
...     outputs = model(**inputs)

>>> last_hidden_states = outputs.last_hidden_state
>>> list(last_hidden_states.shape)
[1, 640, 8, 8]

MobileViTForImageClassification

class transformers.MobileViTForImageClassification

< >

( config: MobileViTConfig )

Parameters

  • config (MobileViTConfig) — 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.

MobileViT model with an image classification head on top (a linear layer on top of the pooled features), e.g. for ImageNet.

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: typing.Optional[torch.Tensor] = None output_hidden_states: typing.Optional[bool] = None labels: typing.Optional[torch.Tensor] = None return_dict: typing.Optional[bool] = None ) → transformers.modeling_outputs.ImageClassifierOutputWithNoAttention 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 MobileViTFeatureExtractor. See MobileViTFeatureExtractor.call() for details.
  • 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,), optional) — Labels for computing the image classification/regression loss. Indices should be in [0, ..., config.num_labels - 1]. If config.num_labels == 1 a regression loss is computed (Mean-Square loss). If config.num_labels > 1 a classification loss is computed (Cross-Entropy).

A transformers.modeling_outputs.ImageClassifierOutputWithNoAttention 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 (MobileViTConfig) 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)) — Classification (or regression if config.num_labels==1) scores (before SoftMax).
  • 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 stage) of shape (batch_size, num_channels, height, width). Hidden-states (also called feature maps) of the model at the output of each stage.

The MobileViTForImageClassification 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 MobileViTFeatureExtractor, MobileViTForImageClassification
>>> import torch
>>> from datasets import load_dataset

>>> dataset = load_dataset("huggingface/cats-image")
>>> image = dataset["test"]["image"][0]

>>> feature_extractor = MobileViTFeatureExtractor.from_pretrained("apple/mobilevit-small")
>>> model = MobileViTForImageClassification.from_pretrained("apple/mobilevit-small")

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

>>> with torch.no_grad():
...     logits = model(**inputs).logits

>>> # model predicts one of the 1000 ImageNet classes
>>> predicted_label = logits.argmax(-1).item()
>>> print(model.config.id2label[predicted_label])
tabby, tabby cat

MobileViTForSemanticSegmentation

class transformers.MobileViTForSemanticSegmentation

< >

( config: MobileViTConfig )

Parameters

  • config (MobileViTConfig) — 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.

MobileViT model with a semantic segmentation head on top, e.g. for Pascal VOC.

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: typing.Optional[torch.Tensor] = None labels: typing.Optional[torch.Tensor] = None output_hidden_states: typing.Optional[bool] = None return_dict: typing.Optional[bool] = 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 MobileViTFeatureExtractor. See MobileViTFeatureExtractor.call() for details.
  • 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 (MobileViTConfig) 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 MobileViTForSemanticSegmentation 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 MobileViTFeatureExtractor, MobileViTForSemanticSegmentation
>>> 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 = MobileViTFeatureExtractor.from_pretrained("apple/deeplabv3-mobilevit-small")
>>> model = MobileViTForSemanticSegmentation.from_pretrained("apple/deeplabv3-mobilevit-small")

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

>>> with torch.no_grad():
...     outputs = model(**inputs)

>>> # logits are of shape (batch_size, num_labels, height, width)
>>> logits = outputs.logits