Transformers documentation

Video Vision Transformer (ViViT)

Transformers

You are viewing v4.46.2 version. A newer version v4.47.1 is available.

Join the Hugging Face community

and get access to the augmented documentation experience

Collaborate on models, datasets and Spaces

Faster examples with accelerated inference

Switch between documentation themes

to get started

Video Vision Transformer (ViViT)

Overview

The Vivit model was proposed in ViViT: A Video Vision Transformer by Anurag Arnab, Mostafa Dehghani, Georg Heigold, Chen Sun, Mario Lučić, Cordelia Schmid. The paper proposes one of the first successful pure-transformer based set of models for video understanding.

The abstract from the paper is the following:

We present pure-transformer based models for video classification, drawing upon the recent success of such models in image classification. Our model extracts spatio-temporal tokens from the input video, which are then encoded by a series of transformer layers. In order to handle the long sequences of tokens encountered in video, we propose several, efficient variants of our model which factorise the spatial- and temporal-dimensions of the input. Although transformer-based models are known to only be effective when large training datasets are available, we show how we can effectively regularise the model during training and leverage pretrained image models to be able to train on comparatively small datasets. We conduct thorough ablation studies, and achieve state-of-the-art results on multiple video classification benchmarks including Kinetics 400 and 600, Epic Kitchens, Something-Something v2 and Moments in Time, outperforming prior methods based on deep 3D convolutional networks.

This model was contributed by jegormeister. The original code (written in JAX) can be found here.

Using Scaled Dot Product Attention (SDPA)

PyTorch includes a native scaled dot-product attention (SDPA) operator as part of torch.nn.functional. This function encompasses several implementations that can be applied depending on the inputs and the hardware in use. See the official documentation or the GPU Inference page for more information.

SDPA is used by default for torch>=2.1.1 when an implementation is available, but you may also set attn_implementation="sdpa" in from_pretrained() to explicitly request SDPA to be used.

from transformers import VivitModel
model = VivitModel.from_pretrained("google/vivit-b-16x2-kinetics400", attn_implementation="sdpa", torch_dtype=torch.float16)
...

For the best speedups, we recommend loading the model in half-precision (e.g. torch.float16 or torch.bfloat16).

On a local benchmark (A100-40GB, PyTorch 2.3.0, OS Ubuntu 22.04) with float32 and google/vivit-b-16x2-kinetics400 model, we saw the following speedups during inference.

Training

num_training_steps	batch_size	is cuda	Speedup (%)	Eager peak mem (MB)	sdpa peak mem (MB)	Mem saving (%)
100	1	True	7.122	2575.28	5932.54	130.364

Inference

num_batches	batch_size	is cuda	is half	Speedup (%)	Mem eager (MB)	Mem BT (MB)	Mem saved (%)
20	1	True	False	15.422	715.807	317.079	125.75
20	2	True	False	17.146	1234.75	447.175	176.122
20	4	True	False	18.093	2275.82	709.864	220.6
20	8	True	False	19.284	4358.19	1233.24	253.393

VivitConfig

class transformers.VivitConfig

< source >

( image_size = 224 num_frames = 32 tubelet_size = [2, 16, 16] num_channels = 3 hidden_size = 768 num_hidden_layers = 12 num_attention_heads = 12 intermediate_size = 3072 hidden_act = 'gelu_fast' hidden_dropout_prob = 0.0 attention_probs_dropout_prob = 0.0 initializer_range = 0.02 layer_norm_eps = 1e-06 qkv_bias = True **kwargs )

Parameters

image_size (int, optional, defaults to 224) — The size (resolution) of each image.
num_frames (int, optional, defaults to 32) — The number of frames in each video.
tubelet_size (List[int], optional, defaults to [2, 16, 16]) — The size (resolution) of each tubelet.
num_channels (int, optional, defaults to 3) — The number of input channels.
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_fast") — The non-linear activation function (function or string) in the encoder and pooler. If string, "gelu", "relu", "selu", "gelu_fast" and "gelu_new" are supported.
hidden_dropout_prob (float, optional, defaults to 0.0) — The dropout probability for all fully connected layers in the embeddings, encoder, and pooler.
attention_probs_dropout_prob (float, optional, defaults to 0.0) — 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-06) — 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.

This is the configuration class to store the configuration of a VivitModel. It is used to instantiate a ViViT 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 ViViT google/vivit-b-16x2-kinetics400 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 VivitConfig, VivitModel

>>> # Initializing a ViViT google/vivit-b-16x2-kinetics400 style configuration
>>> configuration = VivitConfig()

>>> # Initializing a model (with random weights) from the google/vivit-b-16x2-kinetics400 style configuration
>>> model = VivitModel(configuration)

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

VivitImageProcessor

class transformers.VivitImageProcessor

< source >

( do_resize: bool = True size: Dict = None resample: Resampling = <Resampling.BILINEAR: 2> do_center_crop: bool = True crop_size: Dict = None do_rescale: bool = True rescale_factor: Union = 0.00784313725490196 offset: bool = True do_normalize: bool = True image_mean: Union = None image_std: Union = None **kwargs )

Parameters

do_resize (bool, optional, defaults to True) — Whether to resize the image’s (height, width) dimensions to the specified size. Can be overridden by the do_resize parameter in the preprocess method.
size (Dict[str, int] optional, defaults to {"shortest_edge" -- 256}): Size of the output image after resizing. The shortest edge of the image will be resized to size["shortest_edge"] while maintaining the aspect ratio of the original image. Can be overriden by size in the preprocess method.
resample (PILImageResampling, optional, defaults to Resampling.BILINEAR) — Resampling filter to use if resizing the image. Can be overridden by the resample parameter in the preprocess method.
do_center_crop (bool, optional, defaults to True) — Whether to center crop the image to the specified crop_size. Can be overridden by the do_center_crop parameter in the preprocess method.
crop_size (Dict[str, int], optional, defaults to {"height" -- 224, "width": 224}): Size of the image after applying the center crop. Can be overridden by the crop_size parameter in the preprocess method.
do_rescale (bool, optional, defaults to True) — Whether to rescale the image by the specified scale rescale_factor. Can be overridden by the do_rescale parameter in the preprocess method.
rescale_factor (int or float, optional, defaults to 1/127.5) — Defines the scale factor to use if rescaling the image. Can be overridden by the rescale_factor parameter in the preprocess method.
offset (bool, optional, defaults to True) — Whether to scale the image in both negative and positive directions. Can be overriden by the offset in the preprocess method.
do_normalize (bool, optional, defaults to True) — Whether to normalize the image. Can be overridden by the do_normalize parameter in the preprocess method.
image_mean (float or List[float], optional, defaults to IMAGENET_STANDARD_MEAN) — Mean to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the image_mean parameter in the preprocess method.
image_std (float or List[float], optional, defaults to IMAGENET_STANDARD_STD) — Standard deviation to use if normalizing the image. This is a float or list of floats the length of the number of channels in the image. Can be overridden by the image_std parameter in the preprocess method.

Constructs a Vivit image processor.

preprocess

< source >

( videos: Union do_resize: bool = None size: Dict = None resample: Resampling = None do_center_crop: bool = None crop_size: Dict = None do_rescale: bool = None rescale_factor: float = None offset: bool = None do_normalize: bool = None image_mean: Union = None image_std: Union = None return_tensors: Union = None data_format: ChannelDimension = <ChannelDimension.FIRST: 'channels_first'> input_data_format: Union = None )

Parameters

videos (ImageInput) — Video frames to preprocess. Expects a single or batch of video frames with pixel values ranging from 0 to 255. If passing in frames with pixel values between 0 and 1, set do_rescale=False.
do_resize (bool, optional, defaults to self.do_resize) — Whether to resize the image.
size (Dict[str, int], optional, defaults to self.size) — Size of the image after applying resize.
resample (PILImageResampling, optional, defaults to self.resample) — Resampling filter to use if resizing the image. This can be one of the enum PILImageResampling, Only has an effect if do_resize is set to True.
do_center_crop (bool, optional, defaults to self.do_centre_crop) — Whether to centre crop the image.
crop_size (Dict[str, int], optional, defaults to self.crop_size) — Size of the image after applying the centre crop.
do_rescale (bool, optional, defaults to self.do_rescale) — Whether to rescale the image values between [-1 - 1] if offset is True, [0, 1] otherwise.
rescale_factor (float, optional, defaults to self.rescale_factor) — Rescale factor to rescale the image by if do_rescale is set to True.
offset (bool, optional, defaults to self.offset) — Whether to scale the image in both negative and positive directions.
do_normalize (bool, optional, defaults to self.do_normalize) — Whether to normalize the image.
image_mean (float or List[float], optional, defaults to self.image_mean) — Image mean.
image_std (float or List[float], optional, defaults to self.image_std) — Image standard deviation.
return_tensors (str or TensorType, optional) — The type of tensors to return. Can be one of:
- Unset: Return a list of np.ndarray.
- TensorType.TENSORFLOW or 'tf': Return a batch of type tf.Tensor.
- TensorType.PYTORCH or 'pt': Return a batch of type torch.Tensor.
- TensorType.NUMPY or 'np': Return a batch of type np.ndarray.
- TensorType.JAX or 'jax': Return a batch of type jax.numpy.ndarray.
data_format (ChannelDimension or str, optional, defaults to ChannelDimension.FIRST) — The channel dimension format for the output image. Can be one of:
- ChannelDimension.FIRST: image in (num_channels, height, width) format.
- ChannelDimension.LAST: image in (height, width, num_channels) format.
- Unset: Use the inferred channel dimension format of the input image.
input_data_format (ChannelDimension or str, optional) — The channel dimension format for the input image. If unset, the channel dimension format is inferred from the input image. Can be one of:
- "channels_first" or ChannelDimension.FIRST: image in (num_channels, height, width) format.
- "channels_last" or ChannelDimension.LAST: image in (height, width, num_channels) format.
- "none" or ChannelDimension.NONE: image in (height, width) format.

Preprocess an image or batch of images.

VivitModel

class transformers.VivitModel

< source >

( config add_pooling_layer = True )

Parameters

config (VivitConfig) — 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 ViViT Transformer 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

< source >

( pixel_values: Optional = None head_mask: Optional = None output_attentions: Optional = None output_hidden_states: Optional = None interpolate_pos_encoding: bool = False return_dict: Optional = None ) → transformers.modeling_outputs.BaseModelOutputWithPooling or tuple(torch.FloatTensor)

Parameters

pixel_values (torch.FloatTensor of shape (batch_size, num_frames, num_channels, height, width)) — Pixel values. Pixel values can be obtained using VivitImageProcessor. See VivitImageProcessor.preprocess() 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]:
- 1 indicates the head is not masked,
- 0 indicates the head is masked.
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.
interpolate_pos_encoding (bool, optional, False) — Whether to interpolate the pre-trained position encodings.
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 (VivitConfig) 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 VivitModel 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:

>>> import av
>>> import numpy as np

>>> from transformers import VivitImageProcessor, VivitModel
>>> from huggingface_hub import hf_hub_download

>>> np.random.seed(0)


>>> def read_video_pyav(container, indices):
...     '''
...     Decode the video with PyAV decoder.
...     Args:
...         container (`av.container.input.InputContainer`): PyAV container.
...         indices (`List[int]`): List of frame indices to decode.
...     Returns:
...         result (np.ndarray): np array of decoded frames of shape (num_frames, height, width, 3).
...     '''
...     frames = []
...     container.seek(0)
...     start_index = indices[0]
...     end_index = indices[-1]
...     for i, frame in enumerate(container.decode(video=0)):
...         if i > end_index:
...             break
...         if i >= start_index and i in indices:
...             frames.append(frame)
...     return np.stack([x.to_ndarray(format="rgb24") for x in frames])


>>> def sample_frame_indices(clip_len, frame_sample_rate, seg_len):
...     '''
...     Sample a given number of frame indices from the video.
...     Args:
...         clip_len (`int`): Total number of frames to sample.
...         frame_sample_rate (`int`): Sample every n-th frame.
...         seg_len (`int`): Maximum allowed index of sample's last frame.
...     Returns:
...         indices (`List[int]`): List of sampled frame indices
...     '''
...     converted_len = int(clip_len * frame_sample_rate)
...     end_idx = np.random.randint(converted_len, seg_len)
...     start_idx = end_idx - converted_len
...     indices = np.linspace(start_idx, end_idx, num=clip_len)
...     indices = np.clip(indices, start_idx, end_idx - 1).astype(np.int64)
...     return indices


>>> # video clip consists of 300 frames (10 seconds at 30 FPS)
>>> file_path = hf_hub_download(
...     repo_id="nielsr/video-demo", filename="eating_spaghetti.mp4", repo_type="dataset"
... )
>>> container = av.open(file_path)

>>> # sample 32 frames
>>> indices = sample_frame_indices(clip_len=32, frame_sample_rate=1, seg_len=container.streams.video[0].frames)
>>> video = read_video_pyav(container=container, indices=indices)

>>> image_processor = VivitImageProcessor.from_pretrained("google/vivit-b-16x2-kinetics400")
>>> model = VivitModel.from_pretrained("google/vivit-b-16x2-kinetics400")

>>> # prepare video for the model
>>> inputs = image_processor(list(video), return_tensors="pt")

>>> # forward pass
>>> outputs = model(**inputs)
>>> last_hidden_states = outputs.last_hidden_state
>>> list(last_hidden_states.shape)
[1, 3137, 768]

VivitForVideoClassification

class transformers.VivitForVideoClassification

< source >

( config )

Parameters

config (VivitConfig) — 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.

ViViT Transformer model with a video classification head on top (a linear layer on top of the final hidden state of the [CLS] token) e.g. for Kinetics-400.

Note that it’s possible to fine-tune ViT on higher resolution images than the ones it has been trained on, by setting interpolate_pos_encoding to True in the forward of the model. This will interpolate the pre-trained position embeddings to the higher resolution.

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

< source >

( pixel_values: Optional = None head_mask: Optional = None labels: Optional = None output_attentions: Optional = None output_hidden_states: Optional = None interpolate_pos_encoding: bool = False return_dict: Optional = None ) → transformers.modeling_outputs.ImageClassifierOutput or tuple(torch.FloatTensor)

Parameters

pixel_values (torch.FloatTensor of shape (batch_size, num_frames, num_channels, height, width)) — Pixel values. Pixel values can be obtained using VivitImageProcessor. See VivitImageProcessor.preprocess() 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]:
- 1 indicates the head is not masked,
- 0 indicates the head is masked.
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.
interpolate_pos_encoding (bool, optional, False) — Whether to interpolate the pre-trained position encodings.
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).

Returns

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

A transformers.modeling_outputs.ImageClassifierOutput 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 (VivitConfig) 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, sequence_length, hidden_size). Hidden-states (also called feature maps) of the model at the output of each stage.
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 VivitForVideoClassification forward method, overrides the __call__ special method.

Examples:

>>> import av
>>> import numpy as np
>>> import torch

>>> from transformers import VivitImageProcessor, VivitForVideoClassification
>>> from huggingface_hub import hf_hub_download

>>> np.random.seed(0)


>>> def read_video_pyav(container, indices):
...     '''
...     Decode the video with PyAV decoder.
...     Args:
...         container (`av.container.input.InputContainer`): PyAV container.
...         indices (`List[int]`): List of frame indices to decode.
...     Returns:
...         result (np.ndarray): np array of decoded frames of shape (num_frames, height, width, 3).
...     '''
...     frames = []
...     container.seek(0)
...     start_index = indices[0]
...     end_index = indices[-1]
...     for i, frame in enumerate(container.decode(video=0)):
...         if i > end_index:
...             break
...         if i >= start_index and i in indices:
...             frames.append(frame)
...     return np.stack([x.to_ndarray(format="rgb24") for x in frames])


>>> def sample_frame_indices(clip_len, frame_sample_rate, seg_len):
...     '''
...     Sample a given number of frame indices from the video.
...     Args:
...         clip_len (`int`): Total number of frames to sample.
...         frame_sample_rate (`int`): Sample every n-th frame.
...         seg_len (`int`): Maximum allowed index of sample's last frame.
...     Returns:
...         indices (`List[int]`): List of sampled frame indices
...     '''
...     converted_len = int(clip_len * frame_sample_rate)
...     end_idx = np.random.randint(converted_len, seg_len)
...     start_idx = end_idx - converted_len
...     indices = np.linspace(start_idx, end_idx, num=clip_len)
...     indices = np.clip(indices, start_idx, end_idx - 1).astype(np.int64)
...     return indices


>>> # video clip consists of 300 frames (10 seconds at 30 FPS)
>>> file_path = hf_hub_download(
...     repo_id="nielsr/video-demo", filename="eating_spaghetti.mp4", repo_type="dataset"
... )
>>> container = av.open(file_path)

>>> # sample 32 frames
>>> indices = sample_frame_indices(clip_len=32, frame_sample_rate=4, seg_len=container.streams.video[0].frames)
>>> video = read_video_pyav(container=container, indices=indices)

>>> image_processor = VivitImageProcessor.from_pretrained("google/vivit-b-16x2-kinetics400")
>>> model = VivitForVideoClassification.from_pretrained("google/vivit-b-16x2-kinetics400")

>>> inputs = image_processor(list(video), return_tensors="pt")

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

>>> # model predicts one of the 400 Kinetics-400 classes
>>> predicted_label = logits.argmax(-1).item()
>>> print(model.config.id2label[predicted_label])
LABEL_116

< > Update on GitHub

←VideoMAE ALIGN→