LayoutLMv3 | Object Detection & Huggingface Transformers #800
Comments
|
Currently, LayoutLMv3 in Transformers does not support object detection (see @NielsRogge's reply below).
You may refer to the code related to You can also use DiT for the object detection task. DiT and LayoutLMv3 have the same architecture but different backbone models regarding the object detection task. |
Ahhh thanks! This was really helpful. I appreciate your time. I am new to the Detectron2 API and Object Detection. I note that the Detectron2 config includes:
Do both Object Detection tasks for DiT and LayoutLMv3 make use of the embedding vectors of the text (via OCR) as well as the visual context from the when predicting BBoxes? Or is that just in the Huggingfaces transformers implementation? I can't seem to locate the OCR pre-processing code. |
|
|
Ahh ok. You confirmed my suspicions. Some quick follow ups:
|
|
Interesting... I think I will try use the HF YOLOS ObjectDetection prediction head in transformers with LayoutLMv3Model as the base instead of the ViT.
Nice! |
|
Hi Jordan, It's really an exciting idea. In fact I am also working on a similar use case. How was the approach? |
|
Here is the code I used to define a model that used LayoutLMv3 and YOLOS Object Detection head. I was able to train a model using this; however, I found that the 512 token cap been exceeded and i think truncation was affecting model performance. However, I could be wrong. If you end up using below, please let me know how your results go and what your findings are. Happy to collaborate with you on this further. import torch
import torch.utils.checkpoint
from torch import Tensor, nn
from dataclasses import dataclass
from typing import Optional, Dict, Tuple, List
from transformers import LayoutLMv3Model, LayoutLMv3PreTrainedModel, LayoutLMv3Config
from transformers.utils import ModelOutput, is_scipy_available, is_vision_available, requires_backends
from configuration import LayoutLMv3ForObjectDetectionConfig
if is_scipy_available():
from scipy.optimize import linear_sum_assignment
if is_vision_available():
from transformers.models.detr.feature_extraction_detr import center_to_corners_format
@dataclass
class LayoutLMv3ObjectDetectionOutput(ModelOutput):
"""
Output type of [`LayoutLMv3ForObjectDetection`].
Args:
loss (`torch.FloatTensor` of shape `(1,)`, *optional*, returned when `labels` are provided)):
Total loss as a linear combination of a negative log-likehood (cross-entropy) for class prediction and a
bounding box loss. The latter is defined as a linear combination of the L1 loss and the generalized
scale-invariant IoU loss.
loss_dict (`Dict`, *optional*):
A dictionary containing the individual losses. Useful for logging.
logits (`torch.FloatTensor` of shape `(batch_size, num_queries, num_classes + 1)`):
Classification logits (including no-object) for all queries.
pred_boxes (`torch.FloatTensor` of shape `(batch_size, num_queries, 4)`):
Normalized boxes coordinates for all queries, represented as (center_x, center_y, width, height). These
values are normalized in [0, 1], relative to the size of each individual image in the batch (disregarding
possible padding). You can use [`~DetrFeatureExtractor.post_process`] to retrieve the unnormalized bounding
boxes.
auxiliary_outputs (`list[Dict]`, *optional*):
Optional, only returned when auxilary losses are activated (i.e. `config.auxiliary_loss` is set to `True`)
and labels are provided. It is a list of dictionaries containing the two above keys (`logits` and
`pred_boxes`) for each decoder layer.
last_hidden_state (`torch.FloatTensor` of shape `(batch_size, sequence_length, hidden_size)`, *optional*):
Sequence of hidden-states at the output of the last layer of the decoder of the model.
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.
"""
loss: Optional[torch.FloatTensor] = None
loss_dict: Optional[Dict] = None
logits: torch.FloatTensor = None
pred_boxes: torch.FloatTensor = None
auxiliary_outputs: Optional[List[Dict]] = None
last_hidden_state: Optional[torch.FloatTensor] = None
hidden_states: Optional[Tuple[torch.FloatTensor]] = None
attentions: Optional[Tuple[torch.FloatTensor]] = None
class LayoutLMv3ForObjectDetection(LayoutLMv3PreTrainedModel):
def __init__(self, config: LayoutLMv3ForObjectDetectionConfig):
super().__init__(config)
# LayoutLMv3 (ViT) encoder model
self.vit = LayoutLMv3Model(config)
# Object detection heads
# We add one for the "no object" class
self.class_labels_classifier = LayoutLMv3MLPPredictionHead(
input_dim=config.hidden_size, hidden_dim=config.hidden_size, output_dim=config.num_labels + 1, num_layers=3
)
self.bbox_predictor = LayoutLMv3MLPPredictionHead(
input_dim=config.hidden_size, hidden_dim=config.hidden_size, output_dim=4, num_layers=3
)
# Initialize weights and apply final processing
self.post_init()
# taken from https://github.com/facebookresearch/detr/blob/master/models/detr.py
@torch.jit.unused
def _set_aux_loss(self, outputs_class, outputs_coord):
# this is a workaround to make torchscript happy, as torchscript
# doesn't support dictionary with non-homogeneous values, such
# as a dict having both a Tensor and a list.
return [{"logits": a, "pred_boxes": b} for a, b in zip(outputs_class[:-1], outputs_coord[:-1])]
def forward(
self,
input_ids=None,
bbox=None,
attention_mask=None,
token_type_ids=None,
position_ids=None,
head_mask=None,
inputs_embeds=None,
pixel_values=None,
labels=None,
output_attentions=None,
output_hidden_states=None,
return_dict=None,
):
r"""
labels (`List[Dict]` of len `(batch_size,)`, *optional*):
Labels for computing the bipartite matching loss. List of dicts, each dictionary containing at least the
following 2 keys: `'class_labels'` and `'boxes'` (the class labels and bounding boxes of an image in the
batch respectively). The class labels themselves should be a `torch.LongTensor` of len `(number of bounding
boxes in the image,)` and the boxes a `torch.FloatTensor` of shape `(number of bounding boxes in the image,
4)`.
Returns:
Examples:
```python
>>> from transformers import LayoutLMv3FeatureExtractor, LayoutLMv3ForObjectDetection
>>> 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 = LayoutLMv3FeatureExtractor.from_pretrained("hustvl/LayoutLMv3-small")
>>> model = LayoutLMv3ForObjectDetection.from_pretrained("hustvl/LayoutLMv3-small")
>>> inputs = feature_extractor(images=image, return_tensors="pt")
>>> outputs = model(**inputs)
>>> # model predicts bounding boxes and corresponding COCO classes
>>> logits = outputs.logits
>>> bboxes = outputs.pred_boxes
```"""
return_dict = return_dict if return_dict is not None else self.config.use_return_dict
# First, sent images through LayoutLMv3 base model to obtain hidden states
outputs = self.vit(
input_ids=input_ids,
bbox=bbox,
attention_mask=attention_mask,
token_type_ids=token_type_ids,
position_ids=position_ids,
head_mask=head_mask,
inputs_embeds=inputs_embeds,
pixel_values=pixel_values,
output_attentions=output_attentions,
output_hidden_states=output_hidden_states,
return_dict=return_dict,
)
sequence_output = outputs[0]
# Take the final hidden states of the detection tokens
sequence_output = sequence_output[:, -self.config.num_detection_tokens :, :]
# Class logits + predicted bounding boxes
logits = self.class_labels_classifier(sequence_output)
pred_boxes = self.bbox_predictor(sequence_output).sigmoid()
loss, loss_dict, auxiliary_outputs = None, None, None
if labels is not None:
# First: create the matcher
matcher = LayoutLMv3HungarianMatcher(
class_cost=self.config.class_cost, bbox_cost=self.config.bbox_cost, giou_cost=self.config.giou_cost
)
# Second: create the criterion
losses = ["labels", "boxes", "cardinality"]
criterion = LayoutLMv3Loss(
matcher=matcher,
num_classes=self.config.num_labels,
eos_coef=self.config.eos_coefficient,
losses=losses,
)
criterion.to(self.device)
# Third: compute the losses, based on outputs and labels
outputs_loss = {}
outputs_loss["logits"] = logits
outputs_loss["pred_boxes"] = pred_boxes
if self.config.auxiliary_loss:
intermediate = outputs.intermediate_hidden_states if return_dict else outputs[4]
outputs_class = self.class_labels_classifier(intermediate)
outputs_coord = self.bbox_predictor(intermediate).sigmoid()
auxiliary_outputs = self._set_aux_loss(outputs_class, outputs_coord)
outputs_loss["auxiliary_outputs"] = auxiliary_outputs
loss_dict = criterion(outputs_loss, labels)
# Fourth: compute total loss, as a weighted sum of the various losses
weight_dict = {"loss_ce": 1, "loss_bbox": self.config.bbox_loss_coefficient}
weight_dict["loss_giou"] = self.config.giou_loss_coefficient
if self.config.auxiliary_loss:
aux_weight_dict = {}
for i in range(self.config.decoder_layers - 1):
aux_weight_dict.update({k + f"_{i}": v for k, v in weight_dict.items()})
weight_dict.update(aux_weight_dict)
loss = sum(loss_dict[k] * weight_dict[k] for k in loss_dict.keys() if k in weight_dict)
if not return_dict:
if auxiliary_outputs is not None:
output = (logits, pred_boxes) + auxiliary_outputs + outputs
else:
output = (logits, pred_boxes) + outputs
return ((loss, loss_dict) + output) if loss is not None else output
return LayoutLMv3ObjectDetectionOutput(
loss=loss,
loss_dict=loss_dict,
logits=logits,
pred_boxes=pred_boxes,
auxiliary_outputs=auxiliary_outputs,
last_hidden_state=outputs.last_hidden_state,
hidden_states=outputs.hidden_states,
attentions=outputs.attentions,
)
# --- LOSS
# Copied from transformers.models.detr.modeling_detr.dice_loss
def dice_loss(inputs, targets, num_boxes):
"""
Compute the DICE loss, similar to generalized IOU for masks
Args:
inputs: A float tensor of arbitrary shape.
The predictions for each example.
targets: A float tensor with the same shape as inputs. Stores the binary
classification label for each element in inputs (0 for the negative class and 1 for the positive
class).
"""
inputs = inputs.sigmoid()
inputs = inputs.flatten(1)
numerator = 2 * (inputs * targets).sum(1)
denominator = inputs.sum(-1) + targets.sum(-1)
loss = 1 - (numerator + 1) / (denominator + 1)
return loss.sum() / num_boxes
# Copied from transformers.models.detr.modeling_detr.sigmoid_focal_loss
def sigmoid_focal_loss(inputs, targets, num_boxes, alpha: float = 0.25, gamma: float = 2):
"""
Loss used in RetinaNet for dense detection: https://arxiv.org/abs/1708.02002.
Args:
inputs: A float tensor of arbitrary shape.
The predictions for each example.
targets: A float tensor with the same shape as inputs. Stores the binary
classification label for each element in inputs (0 for the negative class and 1 for the positive
class).
alpha: (optional) Weighting factor in range (0,1) to balance
positive vs negative examples. Default = -1 (no weighting).
gamma: Exponent of the modulating factor (1 - p_t) to
balance easy vs hard examples.
Returns:
Loss tensor
"""
prob = inputs.sigmoid()
ce_loss = nn.functional.binary_cross_entropy_with_logits(inputs, targets, reduction="none")
p_t = prob * targets + (1 - prob) * (1 - targets)
loss = ce_loss * ((1 - p_t) ** gamma)
if alpha >= 0:
alpha_t = alpha * targets + (1 - alpha) * (1 - targets)
loss = alpha_t * loss
return loss.mean(1).sum() / num_boxes
# Copied from transformers.models.detr.modeling_detr.DetrLoss with Detr->LayoutLMv3
class LayoutLMv3Loss(nn.Module):
"""
This class computes the losses for LayoutLMv3ForObjectDetection/LayoutLMv3ForSegmentation. The process happens in two steps:
1) we compute hungarian assignment between ground truth boxes and the outputs of the model 2) we supervise each
pair of matched ground-truth / prediction (supervise class and box).
A note on the `num_classes` argument (copied from original repo in detr.py): "the naming of the `num_classes`
parameter of the criterion is somewhat misleading. It indeed corresponds to `max_obj_id` + 1, where `max_obj_id` is
the maximum id for a class in your dataset. For example, COCO has a `max_obj_id` of 90, so we pass `num_classes` to
be 91. As another example, for a dataset that has a single class with `id` 1, you should pass `num_classes` to be 2
(`max_obj_id` + 1). For more details on this, check the following discussion
https://github.com/facebookresearch/detr/issues/108#issuecomment-650269223"
Args:
matcher (`LayoutLMv3HungarianMatcher`):
Module able to compute a matching between targets and proposals.
num_classes (`int`):
Number of object categories, omitting the special no-object category.
eos_coef (`float`):
Relative classification weight applied to the no-object category.
losses (`List[str]`):
List of all the losses to be applied. See `get_loss` for a list of all available losses.
"""
def __init__(self, matcher, num_classes, eos_coef, losses):
super().__init__()
self.matcher = matcher
self.num_classes = num_classes
self.eos_coef = eos_coef
self.losses = losses
empty_weight = torch.ones(self.num_classes + 1)
empty_weight[-1] = self.eos_coef
self.register_buffer("empty_weight", empty_weight)
# removed logging parameter, which was part of the original implementation
def loss_labels(self, outputs, targets, indices, num_boxes):
"""
Classification loss (NLL) targets dicts must contain the key "class_labels" containing a tensor of dim
[nb_target_boxes]
"""
if "logits" not in outputs:
raise KeyError("No logits were found in the outputs")
src_logits = outputs["logits"]
idx = self._get_src_permutation_idx(indices)
target_classes_o = torch.cat([t["class_labels"][J] for t, (_, J) in zip(targets, indices)])
target_classes = torch.full(
src_logits.shape[:2], self.num_classes, dtype=torch.int64, device=src_logits.device
)
target_classes[idx] = target_classes_o
loss_ce = nn.functional.cross_entropy(src_logits.transpose(1, 2), target_classes, self.empty_weight)
losses = {"loss_ce": loss_ce}
return losses
@torch.no_grad()
def loss_cardinality(self, outputs, targets, indices, num_boxes):
"""
Compute the cardinality error, i.e. the absolute error in the number of predicted non-empty boxes.
This is not really a loss, it is intended for logging purposes only. It doesn't propagate gradients.
"""
logits = outputs["logits"]
device = logits.device
tgt_lengths = torch.as_tensor([len(v["class_labels"]) for v in targets], device=device)
# Count the number of predictions that are NOT "no-object" (which is the last class)
card_pred = (logits.argmax(-1) != logits.shape[-1] - 1).sum(1)
card_err = nn.functional.l1_loss(card_pred.float(), tgt_lengths.float())
losses = {"cardinality_error": card_err}
return losses
def loss_boxes(self, outputs, targets, indices, num_boxes):
"""
Compute the losses related to the bounding boxes, the L1 regression loss and the GIoU loss.
Targets dicts must contain the key "boxes" containing a tensor of dim [nb_target_boxes, 4]. The target boxes
are expected in format (center_x, center_y, w, h), normalized by the image size.
"""
if "pred_boxes" not in outputs:
raise KeyError("No predicted boxes found in outputs")
idx = self._get_src_permutation_idx(indices)
src_boxes = outputs["pred_boxes"][idx]
target_boxes = torch.cat([t["boxes"][i] for t, (_, i) in zip(targets, indices)], dim=0)
loss_bbox = nn.functional.l1_loss(src_boxes, target_boxes, reduction="none")
losses = {}
losses["loss_bbox"] = loss_bbox.sum() / num_boxes
loss_giou = 1 - torch.diag(
generalized_box_iou(center_to_corners_format(src_boxes), center_to_corners_format(target_boxes))
)
losses["loss_giou"] = loss_giou.sum() / num_boxes
return losses
def loss_masks(self, outputs, targets, indices, num_boxes):
"""
Compute the losses related to the masks: the focal loss and the dice loss.
Targets dicts must contain the key "masks" containing a tensor of dim [nb_target_boxes, h, w].
"""
if "pred_masks" not in outputs:
raise KeyError("No predicted masks found in outputs")
src_idx = self._get_src_permutation_idx(indices)
tgt_idx = self._get_tgt_permutation_idx(indices)
src_masks = outputs["pred_masks"]
src_masks = src_masks[src_idx]
masks = [t["masks"] for t in targets]
# TODO use valid to mask invalid areas due to padding in loss
target_masks, valid = nested_tensor_from_tensor_list(masks).decompose()
target_masks = target_masks.to(src_masks)
target_masks = target_masks[tgt_idx]
# upsample predictions to the target size
src_masks = nn.functional.interpolate(
src_masks[:, None], size=target_masks.shape[-2:], mode="bilinear", align_corners=False
)
src_masks = src_masks[:, 0].flatten(1)
target_masks = target_masks.flatten(1)
target_masks = target_masks.view(src_masks.shape)
losses = {
"loss_mask": sigmoid_focal_loss(src_masks, target_masks, num_boxes),
"loss_dice": dice_loss(src_masks, target_masks, num_boxes),
}
return losses
def _get_src_permutation_idx(self, indices):
# permute predictions following indices
batch_idx = torch.cat([torch.full_like(src, i) for i, (src, _) in enumerate(indices)])
src_idx = torch.cat([src for (src, _) in indices])
return batch_idx, src_idx
def _get_tgt_permutation_idx(self, indices):
# permute targets following indices
batch_idx = torch.cat([torch.full_like(tgt, i) for i, (_, tgt) in enumerate(indices)])
tgt_idx = torch.cat([tgt for (_, tgt) in indices])
return batch_idx, tgt_idx
def get_loss(self, loss, outputs, targets, indices, num_boxes):
loss_map = {
"labels": self.loss_labels,
"cardinality": self.loss_cardinality,
"boxes": self.loss_boxes,
"masks": self.loss_masks,
}
if loss not in loss_map:
raise ValueError(f"Loss {loss} not supported")
return loss_map[loss](outputs, targets, indices, num_boxes)
def forward(self, outputs, targets):
"""
This performs the loss computation.
Args:
outputs (`dict`, *optional*):
Dictionary of tensors, see the output specification of the model for the format.
targets (`List[dict]`, *optional*):
List of dicts, such that len(targets) == batch_size. The expected keys in each dict depends on the
losses applied, see each loss' doc.
"""
outputs_without_aux = {k: v for k, v in outputs.items() if k != "auxiliary_outputs"}
# Retrieve the matching between the outputs of the last layer and the targets
indices = self.matcher(outputs_without_aux, targets)
# Compute the average number of target boxes across all nodes, for normalization purposes
num_boxes = sum(len(t["class_labels"]) for t in targets)
num_boxes = torch.as_tensor([num_boxes], dtype=torch.float, device=next(iter(outputs.values())).device)
# (Niels): comment out function below, distributed training to be added
# if is_dist_avail_and_initialized():
# torch.distributed.all_reduce(num_boxes)
# (Niels) in original implementation, num_boxes is divided by get_world_size()
num_boxes = torch.clamp(num_boxes, min=1).item()
# Compute all the requested losses
losses = {}
for loss in self.losses:
losses.update(self.get_loss(loss, outputs, targets, indices, num_boxes))
# In case of auxiliary losses, we repeat this process with the output of each intermediate layer.
if "auxiliary_outputs" in outputs:
for i, auxiliary_outputs in enumerate(outputs["auxiliary_outputs"]):
indices = self.matcher(auxiliary_outputs, targets)
for loss in self.losses:
if loss == "masks":
# Intermediate masks losses are too costly to compute, we ignore them.
continue
l_dict = self.get_loss(loss, auxiliary_outputs, targets, indices, num_boxes)
l_dict = {k + f"_{i}": v for k, v in l_dict.items()}
losses.update(l_dict)
return losses
# Copied from transformers.models.detr.modeling_detr.DetrMLPPredictionHead with Detr->LayoutLMv3
class LayoutLMv3MLPPredictionHead(nn.Module):
"""
Very simple multi-layer perceptron (MLP, also called FFN), used to predict the normalized center coordinates,
height and width of a bounding box w.r.t. an image.
Copied from https://github.com/facebookresearch/detr/blob/master/models/detr.py
"""
def __init__(self, input_dim, hidden_dim, output_dim, num_layers):
super().__init__()
self.num_layers = num_layers
h = [hidden_dim] * (num_layers - 1)
self.layers = nn.ModuleList(nn.Linear(n, k) for n, k in zip([input_dim] + h, h + [output_dim]))
def forward(self, x):
for i, layer in enumerate(self.layers):
x = nn.functional.relu(layer(x)) if i < self.num_layers - 1 else layer(x)
return x
# Copied from transformers.models.detr.modeling_detr.DetrHungarianMatcher with Detr->LayoutLMv3
class LayoutLMv3HungarianMatcher(nn.Module):
"""
This class computes an assignment between the targets and the predictions of the network.
For efficiency reasons, the targets don't include the no_object. Because of this, in general, there are more
predictions than targets. In this case, we do a 1-to-1 matching of the best predictions, while the others are
un-matched (and thus treated as non-objects).
Args:
class_cost:
The relative weight of the classification error in the matching cost.
bbox_cost:
The relative weight of the L1 error of the bounding box coordinates in the matching cost.
giou_cost:
The relative weight of the giou loss of the bounding box in the matching cost.
"""
def __init__(self, class_cost: float = 1, bbox_cost: float = 1, giou_cost: float = 1):
super().__init__()
requires_backends(self, ["scipy"])
self.class_cost = class_cost
self.bbox_cost = bbox_cost
self.giou_cost = giou_cost
if class_cost == 0 or bbox_cost == 0 or giou_cost == 0:
raise ValueError("All costs of the Matcher can't be 0")
@torch.no_grad()
def forward(self, outputs, targets):
"""
Args:
outputs (`dict`):
A dictionary that contains at least these entries:
* "logits": Tensor of dim [batch_size, num_queries, num_classes] with the classification logits
* "pred_boxes": Tensor of dim [batch_size, num_queries, 4] with the predicted box coordinates.
targets (`List[dict]`):
A list of targets (len(targets) = batch_size), where each target is a dict containing:
* "class_labels": Tensor of dim [num_target_boxes] (where num_target_boxes is the number of
ground-truth
objects in the target) containing the class labels
* "boxes": Tensor of dim [num_target_boxes, 4] containing the target box coordinates.
Returns:
`List[Tuple]`: A list of size `batch_size`, containing tuples of (index_i, index_j) where:
- index_i is the indices of the selected predictions (in order)
- index_j is the indices of the corresponding selected targets (in order)
For each batch element, it holds: len(index_i) = len(index_j) = min(num_queries, num_target_boxes)
"""
batch_size, num_queries = outputs["logits"].shape[:2]
# We flatten to compute the cost matrices in a batch
out_prob = outputs["logits"].flatten(0, 1).softmax(-1) # [batch_size * num_queries, num_classes]
out_bbox = outputs["pred_boxes"].flatten(0, 1) # [batch_size * num_queries, 4]
# Also concat the target labels and boxes
tgt_ids = torch.cat([v["class_labels"] for v in targets])
tgt_bbox = torch.cat([v["boxes"] for v in targets])
# Compute the classification cost. Contrary to the loss, we don't use the NLL,
# but approximate it in 1 - proba[target class].
# The 1 is a constant that doesn't change the matching, it can be ommitted.
class_cost = -out_prob[:, tgt_ids]
# Compute the L1 cost between boxes
bbox_cost = torch.cdist(out_bbox, tgt_bbox, p=1)
# Compute the giou cost between boxes
giou_cost = -generalized_box_iou(center_to_corners_format(out_bbox), center_to_corners_format(tgt_bbox))
# Final cost matrix
cost_matrix = self.bbox_cost * bbox_cost + self.class_cost * class_cost + self.giou_cost * giou_cost
cost_matrix = cost_matrix.view(batch_size, num_queries, -1).cpu()
sizes = [len(v["boxes"]) for v in targets]
indices = [linear_sum_assignment(c[i]) for i, c in enumerate(cost_matrix.split(sizes, -1))]
return [(torch.as_tensor(i, dtype=torch.int64), torch.as_tensor(j, dtype=torch.int64)) for i, j in indices]
# Copied from transformers.models.detr.modeling_detr._upcast
def _upcast(t: Tensor) -> Tensor:
# Protects from numerical overflows in multiplications by upcasting to the equivalent higher type
if t.is_floating_point():
return t if t.dtype in (torch.float32, torch.float64) else t.float()
else:
return t if t.dtype in (torch.int32, torch.int64) else t.int()
# Copied from transformers.models.detr.modeling_detr.box_area
def box_area(boxes: Tensor) -> Tensor:
"""
Computes the area of a set of bounding boxes, which are specified by its (x1, y1, x2, y2) coordinates.
Args:
boxes (`torch.FloatTensor` of shape `(number_of_boxes, 4)`):
Boxes for which the area will be computed. They are expected to be in (x1, y1, x2, y2) format with `0 <= x1
< x2` and `0 <= y1 < y2`.
Returns:
`torch.FloatTensor`: a tensor containing the area for each box.
"""
boxes = _upcast(boxes)
return (boxes[:, 2] - boxes[:, 0]) * (boxes[:, 3] - boxes[:, 1])
# Copied from transformers.models.detr.modeling_detr.box_iou
def box_iou(boxes1, boxes2):
area1 = box_area(boxes1)
area2 = box_area(boxes2)
left_top = torch.max(boxes1[:, None, :2], boxes2[:, :2]) # [N,M,2]
right_bottom = torch.min(boxes1[:, None, 2:], boxes2[:, 2:]) # [N,M,2]
width_height = (right_bottom - left_top).clamp(min=0) # [N,M,2]
inter = width_height[:, :, 0] * width_height[:, :, 1] # [N,M]
union = area1[:, None] + area2 - inter
iou = inter / union
return iou, union
# Copied from transformers.models.detr.modeling_detr.generalized_box_iou
def generalized_box_iou(boxes1, boxes2):
"""
Generalized IoU from https://giou.stanford.edu/. The boxes should be in [x0, y0, x1, y1] (corner) format.
Returns:
`torch.FloatTensor`: a [N, M] pairwise matrix, where N = len(boxes1) and M = len(boxes2)
"""
# degenerate boxes gives inf / nan results
# so do an early check
assert (boxes1[:, 2:] >= boxes1[:, :2]).all()
assert (boxes2[:, 2:] >= boxes2[:, :2]).all()
iou, union = box_iou(boxes1, boxes2)
lt = torch.min(boxes1[:, None, :2], boxes2[:, :2])
rb = torch.max(boxes1[:, None, 2:], boxes2[:, 2:])
wh = (rb - lt).clamp(min=0) # [N,M,2]
area = wh[:, :, 0] * wh[:, :, 1]
return iou - (area - union) / area
# Copied from transformers.models.detr.modeling_detr._max_by_axis
def _max_by_axis(the_list):
# type: (List[List[int]]) -> List[int]
maxes = the_list[0]
for sublist in the_list[1:]:
for index, item in enumerate(sublist):
maxes[index] = max(maxes[index], item)
return maxes
# Copied from transformers.models.detr.modeling_detr.NestedTensor
class NestedTensor(object):
def __init__(self, tensors, mask: Optional[Tensor]):
self.tensors = tensors
self.mask = mask
def to(self, device):
cast_tensor = self.tensors.to(device)
mask = self.mask
if mask is not None:
cast_mask = mask.to(device)
else:
cast_mask = None
return NestedTensor(cast_tensor, cast_mask)
def decompose(self):
return self.tensors, self.mask
def __repr__(self):
return str(self.tensors)
# Copied from transformers.models.detr.modeling_detr.nested_tensor_from_tensor_list
def nested_tensor_from_tensor_list(tensor_list: List[Tensor]):
if tensor_list[0].ndim == 3:
max_size = _max_by_axis([list(img.shape) for img in tensor_list])
batch_shape = [len(tensor_list)] + max_size
b, c, h, w = batch_shape
dtype = tensor_list[0].dtype
device = tensor_list[0].device
tensor = torch.zeros(batch_shape, dtype=dtype, device=device)
mask = torch.ones((b, h, w), dtype=torch.bool, device=device)
for img, pad_img, m in zip(tensor_list, tensor, mask):
pad_img[: img.shape[0], : img.shape[1], : img.shape[2]].copy_(img)
m[: img.shape[1], : img.shape[2]] = False
else:
raise ValueError("Only 3-dimensional tensors are supported")
return NestedTensor(tensor, mask) |
|
any update on LayoutLMv3 with combination of image + embedding vectors of text ? |
|
Sry Muhammad couldn't get time to take a look at it.
…On Wed, 19 Apr, 2023, 12:42 pm Muhammed Shameem K, ***@***.***> wrote:
any update on LayoutLMv3 with combination of image+ embedding vectors of
text ?
—
Reply to this email directly, view it on GitHub
<#800 (comment)>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ALUOLCVZT4KSW3W6F2OK2Y3XB6F5FANCNFSM54QW2PGQ>
.
You are receiving this because you commented.Message ID:
***@***.***>
|
Hello @jordanparker6 ! |
|
Hi @jordanparker6 , I tried your script for Can you explain to me how did you use |
|
Hey @jordanparker6 Can you share your config for LayoutLMv3ForObjectDetectionConfig please? |
|
The configuration code is below. Let me know how you go. I think the issue I had was the 512 token size. I was then looking at BART for a larger token window kind of how DONUT does it and then got distracted... Using a LLM might be even better as it would be ideal to fit the entire page into the text context. JPMorgan had a paper using the outputs of OCR as input to a LLM without any vision too and it did well. from transformers.configuration_utils import PretrainedConfig
class LayoutLMv3ForObjectDetectionConfig(PretrainedConfig):
r"""
This is the configuration class to store the configuration of a [`LayoutLMv3Model`]. It is used to instantiate an
LayoutLMv3 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 LayoutLMv3
[microsoft/layoutlmv3-base](https://huggingface.co/microsoft/layoutlmv3-base) architecture.
Configuration objects inherit from [`PretrainedConfig`] and can be used to control the model outputs. Read the
documentation from [`PretrainedConfig`] for more information.
Args:
vocab_size (`int`, *optional*, defaults to 50265):
Vocabulary size of the LayoutLMv3 model. Defines the number of different tokens that can be represented by
the `inputs_ids` passed when calling [`LayoutLMv3Model`].
hidden_size (`int`, *optional*, defaults to 768):
Dimension 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):
Dimension 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.
max_position_embeddings (`int`, *optional*, defaults to 512):
The maximum sequence length that this model might ever be used with. Typically set this to something large
just in case (e.g., 512 or 1024 or 2048).
type_vocab_size (`int`, *optional*, defaults to 2):
The vocabulary size of the `token_type_ids` passed when calling [`LayoutLMv3Model`].
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.
max_2d_position_embeddings (`int`, *optional*, defaults to 1024):
The maximum value that the 2D position embedding might ever be used with. Typically set this to something
large just in case (e.g., 1024).
coordinate_size (`int`, *optional*, defaults to `128`):
Dimension of the coordinate embeddings.
shape_size (`int`, *optional*, defaults to `128`):
Dimension of the width and height embeddings.
has_relative_attention_bias (`bool`, *optional*, defaults to `True`):
Whether or not to use a relative attention bias in the self-attention mechanism.
rel_pos_bins (`int`, *optional*, defaults to 32):
The number of relative position bins to be used in the self-attention mechanism.
max_rel_pos (`int`, *optional*, defaults to 128):
The maximum number of relative positions to be used in the self-attention mechanism.
max_rel_2d_pos (`int`, *optional*, defaults to 256):
The maximum number of relative 2D positions in the self-attention mechanism.
rel_2d_pos_bins (`int`, *optional*, defaults to 64):
The number of 2D relative position bins in the self-attention mechanism.
has_spatial_attention_bias (`bool`, *optional*, defaults to `True`):
Whether or not to use a spatial attention bias in the self-attention mechanism.
visual_embed (`bool`, *optional*, defaults to `True`):
Whether or not to add patch embeddings.
input_size (`int`, *optional*, defaults to `224`):
The size (resolution) of the images.
num_channels (`int`, *optional*, defaults to `3`):
The number of channels of the images.
patch_size (`int`, *optional*, defaults to `16`)
The size (resolution) of the patches.
classifier_dropout (`float`, *optional*):
The dropout ratio for the classification head.
num_detection_tokens (`int`, *optional*, defaults to `100`):
The number of detection tokens.
Example:
```python
>>> from transformers import LayoutLMv3Model, LayoutLMv3Config
>>> # Initializing a LayoutLMv3 microsoft/layoutlmv3-base style configuration
>>> configuration = LayoutLMv3Config()
>>> # Initializing a model from the microsoft/layoutlmv3-base style configuration
>>> model = LayoutLMv3Model(configuration)
>>> # Accessing the model configuration
>>> configuration = model.config
```"""
model_type = "layoutlmv3"
def __init__(
self,
vocab_size=50265,
hidden_size=768,
num_hidden_layers=12,
num_attention_heads=12,
intermediate_size=3072,
hidden_act="gelu",
hidden_dropout_prob=0.1,
attention_probs_dropout_prob=0.1,
max_position_embeddings=512,
type_vocab_size=2,
initializer_range=0.02,
layer_norm_eps=1e-5,
pad_token_id=1,
bos_token_id=0,
eos_token_id=2,
max_2d_position_embeddings=1024,
coordinate_size=128,
shape_size=128,
has_relative_attention_bias=True,
rel_pos_bins=32,
max_rel_pos=128,
rel_2d_pos_bins=64,
max_rel_2d_pos=256,
has_spatial_attention_bias=True,
text_embed=True,
visual_embed=True,
input_size=224,
num_channels=3,
patch_size=16,
classifier_dropout=None,
num_detection_tokens=100,
qkv_bias=True,
use_mid_position_embeddings=True,
auxiliary_loss=False,
class_cost=1,
bbox_cost=5,
giou_cost=2,
bbox_loss_coefficient=5,
giou_loss_coefficient=2,
eos_coefficient=0.1,
**kwargs
):
super().__init__(
vocab_size=vocab_size,
hidden_size=hidden_size,
num_hidden_layers=num_hidden_layers,
num_attention_heads=num_attention_heads,
intermediate_size=intermediate_size,
hidden_act=hidden_act,
hidden_dropout_prob=hidden_dropout_prob,
attention_probs_dropout_prob=attention_probs_dropout_prob,
max_position_embeddings=max_position_embeddings,
type_vocab_size=type_vocab_size,
initializer_range=initializer_range,
layer_norm_eps=layer_norm_eps,
pad_token_id=pad_token_id,
bos_token_id=bos_token_id,
eos_token_id=eos_token_id,
**kwargs,
)
self.max_2d_position_embeddings = max_2d_position_embeddings
self.coordinate_size = coordinate_size
self.shape_size = shape_size
self.has_relative_attention_bias = has_relative_attention_bias
self.rel_pos_bins = rel_pos_bins
self.max_rel_pos = max_rel_pos
self.has_spatial_attention_bias = has_spatial_attention_bias
self.rel_2d_pos_bins = rel_2d_pos_bins
self.max_rel_2d_pos = max_rel_2d_pos
self.text_embed = text_embed
self.visual_embed = visual_embed
self.input_size = input_size
self.num_channels = num_channels
self.patch_size = patch_size
self.classifier_dropout = classifier_dropout
self.num_detection_tokens = num_detection_tokens
self.qkv_bias = qkv_bias
self.use_mid_position_embeddings = use_mid_position_embeddings
self.auxiliary_loss = auxiliary_loss
self.class_cost = class_cost
self.bbox_cost = bbox_cost
self.giou_cost = giou_cost
self.bbox_loss_coefficient = bbox_loss_coefficient
self.giou_loss_coefficient = giou_loss_coefficient
self.eos_coefficient = eos_coefficient |
|
Many Thanks @jordanparker6 ! I am trying to use your code with funsd dataset, however I am not sure if I am modelling the no-object class correctly. So with LMV3 you can pad inputs with Pad_token= -100 but, from what I understand, detr (and yolos by extension since it is built on detr) reserve the top label for the no-object class. So if I have 4 classes (0,1,2,3) and pad class -100 (which I assume will be treated as no object-class) - should I even pass the samples with -100 label? Since the max value of all labels is 3, does it mean that -100 will be treated as class nr 1 and class 3 will be treated as no-object? Is there a need to pass examples of no-object class? How did you model your problem? Can you share example of your training data please? A lot of questions! Apologies:) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Is it possible to use LayoutLMv3 for object detection using the Transformers library?
I can see that LayoutLMv3SequenceClassification and LayoutLMv3TokenClassification exist? I am not sure how these would cover object detection.
Or, do we need to use the DIT (leveraging detectron2) code supplied in this repo at https://github.com/microsoft/unilm/tree/master/layoutlmv3/examples/object_detection?
The text was updated successfully, but these errors were encountered: