quoccuongle/Grounded-SAM-2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

support gsam2 image predictor model

Browse Source
This commit is contained in:
rentainhe
2024-08-01 17:05:01 +08:00
parent 72501fecf8
commit 1dacb47840
333 changed files with 24764 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

0
grounding_dino/groundingdino/datasets/__init__.py Normal file
View File

269
grounding_dino/groundingdino/datasets/cocogrounding_eval.py Normal file
View File

@@ -0,0 +1,269 @@
# ------------------------------------------------------------------------
# Grounding DINO. Midified by Shilong Liu.
# url: https://github.com/IDEA-Research/GroundingDINO
# Copyright (c) 2023 IDEA. All Rights Reserved.
# Licensed under the Apache License, Version 2.0 [see LICENSE for details]
# ------------------------------------------------------------------------
# Copyright (c) Aishwarya Kamath & Nicolas Carion. Licensed under the Apache License 2.0. All Rights Reserved
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
"""
COCO evaluator that works in distributed mode.
Mostly copy-paste from https://github.com/pytorch/vision/blob/edfd5a7/references/detection/coco_eval.py
The difference is that there is less copy-pasting from pycocotools
in the end of the file, as python3 can suppress prints with contextlib
"""
import contextlib
import copy
import os
import numpy as np
import pycocotools.mask as mask_util
import torch
from pycocotools.coco import COCO
from pycocotools.cocoeval import COCOeval
from groundingdino.util.misc import all_gather
class CocoGroundingEvaluator(object):
def __init__(self, coco_gt, iou_types, useCats=True):
assert isinstance(iou_types, (list, tuple))
coco_gt = copy.deepcopy(coco_gt)
self.coco_gt = coco_gt
self.iou_types = iou_types
self.coco_eval = {}
for iou_type in iou_types:
self.coco_eval[iou_type] = COCOeval(coco_gt, iouType=iou_type)
self.coco_eval[iou_type].useCats = useCats
self.img_ids = []
self.eval_imgs = {k: [] for k in iou_types}
self.useCats = useCats
def update(self, predictions):
img_ids = list(np.unique(list(predictions.keys())))
self.img_ids.extend(img_ids)
for iou_type in self.iou_types:
results = self.prepare(predictions, iou_type)
# suppress pycocotools prints
with open(os.devnull, "w") as devnull:
with contextlib.redirect_stdout(devnull):
coco_dt = COCO.loadRes(self.coco_gt, results) if results else COCO()
coco_eval = self.coco_eval[iou_type]
coco_eval.cocoDt = coco_dt
coco_eval.params.imgIds = list(img_ids)
coco_eval.params.useCats = self.useCats
img_ids, eval_imgs = evaluate(coco_eval)
self.eval_imgs[iou_type].append(eval_imgs)
def synchronize_between_processes(self):
for iou_type in self.iou_types:
self.eval_imgs[iou_type] = np.concatenate(self.eval_imgs[iou_type], 2)
create_common_coco_eval(self.coco_eval[iou_type], self.img_ids, self.eval_imgs[iou_type])
def accumulate(self):
for coco_eval in self.coco_eval.values():
coco_eval.accumulate()
def summarize(self):
for iou_type, coco_eval in self.coco_eval.items():
print("IoU metric: {}".format(iou_type))
coco_eval.summarize()
def prepare(self, predictions, iou_type):
if iou_type == "bbox":
return self.prepare_for_coco_detection(predictions)
elif iou_type == "segm":
return self.prepare_for_coco_segmentation(predictions)
elif iou_type == "keypoints":
return self.prepare_for_coco_keypoint(predictions)
else:
raise ValueError("Unknown iou type {}".format(iou_type))
def prepare_for_coco_detection(self, predictions):
coco_results = []
for original_id, prediction in predictions.items():
if len(prediction) == 0:
continue
boxes = prediction["boxes"]
boxes = convert_to_xywh(boxes).tolist()
scores = prediction["scores"].tolist()
labels = prediction["labels"].tolist()
coco_results.extend(
[
{
"image_id": original_id,
"category_id": labels[k],
"bbox": box,
"score": scores[k],
}
for k, box in enumerate(boxes)
]
)
return coco_results
def prepare_for_coco_segmentation(self, predictions):
coco_results = []
for original_id, prediction in predictions.items():
if len(prediction) == 0:
continue
scores = prediction["scores"]
labels = prediction["labels"]
masks = prediction["masks"]
masks = masks > 0.5
scores = prediction["scores"].tolist()
labels = prediction["labels"].tolist()
rles = [
mask_util.encode(np.array(mask[0, :, :, np.newaxis], dtype=np.uint8, order="F"))[0]
for mask in masks
]
for rle in rles:
rle["counts"] = rle["counts"].decode("utf-8")
coco_results.extend(
[
{
"image_id": original_id,
"category_id": labels[k],
"segmentation": rle,
"score": scores[k],
}
for k, rle in enumerate(rles)
]
)
return coco_results
def prepare_for_coco_keypoint(self, predictions):
coco_results = []
for original_id, prediction in predictions.items():
if len(prediction) == 0:
continue
boxes = prediction["boxes"]
boxes = convert_to_xywh(boxes).tolist()
scores = prediction["scores"].tolist()
labels = prediction["labels"].tolist()
keypoints = prediction["keypoints"]
keypoints = keypoints.flatten(start_dim=1).tolist()
coco_results.extend(
[
{
"image_id": original_id,
"category_id": labels[k],
"keypoints": keypoint,
"score": scores[k],
}
for k, keypoint in enumerate(keypoints)
]
)
return coco_results
def convert_to_xywh(boxes):
xmin, ymin, xmax, ymax = boxes.unbind(1)
return torch.stack((xmin, ymin, xmax - xmin, ymax - ymin), dim=1)
def merge(img_ids, eval_imgs):
all_img_ids = all_gather(img_ids)
all_eval_imgs = all_gather(eval_imgs)
merged_img_ids = []
for p in all_img_ids:
merged_img_ids.extend(p)
merged_eval_imgs = []
for p in all_eval_imgs:
merged_eval_imgs.append(p)
merged_img_ids = np.array(merged_img_ids)
merged_eval_imgs = np.concatenate(merged_eval_imgs, 2)
# keep only unique (and in sorted order) images
merged_img_ids, idx = np.unique(merged_img_ids, return_index=True)
merged_eval_imgs = merged_eval_imgs[..., idx]
return merged_img_ids, merged_eval_imgs
def create_common_coco_eval(coco_eval, img_ids, eval_imgs):
img_ids, eval_imgs = merge(img_ids, eval_imgs)
img_ids = list(img_ids)
eval_imgs = list(eval_imgs.flatten())
coco_eval.evalImgs = eval_imgs
coco_eval.params.imgIds = img_ids
coco_eval._paramsEval = copy.deepcopy(coco_eval.params)
#################################################################
# From pycocotools, just removed the prints and fixed
# a Python3 bug about unicode not defined
#################################################################
def evaluate(self):
"""
Run per image evaluation on given images and store results (a list of dict) in self.evalImgs
:return: None
"""
# tic = time.time()
# print('Running per image evaluation...')
p = self.params
# add backward compatibility if useSegm is specified in params
if p.useSegm is not None:
p.iouType = "segm" if p.useSegm == 1 else "bbox"
print("useSegm (deprecated) is not None. Running {} evaluation".format(p.iouType))
# print('Evaluate annotation type *{}*'.format(p.iouType))
p.imgIds = list(np.unique(p.imgIds))
if p.useCats:
p.catIds = list(np.unique(p.catIds))
p.maxDets = sorted(p.maxDets)
self.params = p
self._prepare()
# loop through images, area range, max detection number
catIds = p.catIds if p.useCats else [-1]
if p.iouType == "segm" or p.iouType == "bbox":
computeIoU = self.computeIoU
elif p.iouType == "keypoints":
computeIoU = self.computeOks
self.ious = {
(imgId, catId): computeIoU(imgId, catId)
for imgId in p.imgIds
for catId in catIds}
evaluateImg = self.evaluateImg
maxDet = p.maxDets[-1]
evalImgs = [
evaluateImg(imgId, catId, areaRng, maxDet)
for catId in catIds
for areaRng in p.areaRng
for imgId in p.imgIds
]
# this is NOT in the pycocotools code, but could be done outside
evalImgs = np.asarray(evalImgs).reshape(len(catIds), len(p.areaRng), len(p.imgIds))
self._paramsEval = copy.deepcopy(self.params)
# toc = time.time()
# print('DONE (t={:0.2f}s).'.format(toc-tic))
return p.imgIds, evalImgs
#################################################################
# end of straight copy from pycocotools, just removing the prints
#################################################################

311
grounding_dino/groundingdino/datasets/transforms.py Normal file
View File

@@ -0,0 +1,311 @@
# Copyright (c) Facebook, Inc. and its affiliates. All Rights Reserved
"""
Transforms and data augmentation for both image + bbox.
"""
import os
import random
import PIL
import torch
import torchvision.transforms as T
import torchvision.transforms.functional as F
from grounding_dino.groundingdino.util.box_ops import box_xyxy_to_cxcywh
from grounding_dino.groundingdino.util.misc import interpolate
def crop(image, target, region):
cropped_image = F.crop(image, *region)
target = target.copy()
i, j, h, w = region
# should we do something wrt the original size?
target["size"] = torch.tensor([h, w])
fields = ["labels", "area", "iscrowd", "positive_map"]
if "boxes" in target:
boxes = target["boxes"]
max_size = torch.as_tensor([w, h], dtype=torch.float32)
cropped_boxes = boxes - torch.as_tensor([j, i, j, i])
cropped_boxes = torch.min(cropped_boxes.reshape(-1, 2, 2), max_size)
cropped_boxes = cropped_boxes.clamp(min=0)
area = (cropped_boxes[:, 1, :] - cropped_boxes[:, 0, :]).prod(dim=1)
target["boxes"] = cropped_boxes.reshape(-1, 4)
target["area"] = area
fields.append("boxes")
if "masks" in target:
# FIXME should we update the area here if there are no boxes?
target["masks"] = target["masks"][:, i : i + h, j : j + w]
fields.append("masks")
# remove elements for which the boxes or masks that have zero area
if "boxes" in target or "masks" in target:
# favor boxes selection when defining which elements to keep
# this is compatible with previous implementation
if "boxes" in target:
cropped_boxes = target["boxes"].reshape(-1, 2, 2)
keep = torch.all(cropped_boxes[:, 1, :] > cropped_boxes[:, 0, :], dim=1)
else:
keep = target["masks"].flatten(1).any(1)
for field in fields:
if field in target:
target[field] = target[field][keep]
if os.environ.get("IPDB_SHILONG_DEBUG", None) == "INFO":
# for debug and visualization only.
if "strings_positive" in target:
target["strings_positive"] = [
_i for _i, _j in zip(target["strings_positive"], keep) if _j
]
return cropped_image, target
def hflip(image, target):
flipped_image = F.hflip(image)
w, h = image.size
target = target.copy()
if "boxes" in target:
boxes = target["boxes"]
boxes = boxes[:, [2, 1, 0, 3]] * torch.as_tensor([-1, 1, -1, 1]) + torch.as_tensor(
[w, 0, w, 0]
)
target["boxes"] = boxes
if "masks" in target:
target["masks"] = target["masks"].flip(-1)
return flipped_image, target
def resize(image, target, size, max_size=None):
# size can be min_size (scalar) or (w, h) tuple
def get_size_with_aspect_ratio(image_size, size, max_size=None):
w, h = image_size
if max_size is not None:
min_original_size = float(min((w, h)))
max_original_size = float(max((w, h)))
if max_original_size / min_original_size * size > max_size:
size = int(round(max_size * min_original_size / max_original_size))
if (w <= h and w == size) or (h <= w and h == size):
return (h, w)
if w < h:
ow = size
oh = int(size * h / w)
else:
oh = size
ow = int(size * w / h)
return (oh, ow)
def get_size(image_size, size, max_size=None):
if isinstance(size, (list, tuple)):
return size[::-1]
else:
return get_size_with_aspect_ratio(image_size, size, max_size)
size = get_size(image.size, size, max_size)
rescaled_image = F.resize(image, size)
if target is None:
return rescaled_image, None
ratios = tuple(float(s) / float(s_orig) for s, s_orig in zip(rescaled_image.size, image.size))
ratio_width, ratio_height = ratios
target = target.copy()
if "boxes" in target:
boxes = target["boxes"]
scaled_boxes = boxes * torch.as_tensor(
[ratio_width, ratio_height, ratio_width, ratio_height]
)
target["boxes"] = scaled_boxes
if "area" in target:
area = target["area"]
scaled_area = area * (ratio_width * ratio_height)
target["area"] = scaled_area
h, w = size
target["size"] = torch.tensor([h, w])
if "masks" in target:
target["masks"] = (
interpolate(target["masks"][:, None].float(), size, mode="nearest")[:, 0] > 0.5
)
return rescaled_image, target
def pad(image, target, padding):
# assumes that we only pad on the bottom right corners
padded_image = F.pad(image, (0, 0, padding[0], padding[1]))
if target is None:
return padded_image, None
target = target.copy()
# should we do something wrt the original size?
target["size"] = torch.tensor(padded_image.size[::-1])
if "masks" in target:
target["masks"] = torch.nn.functional.pad(target["masks"], (0, padding[0], 0, padding[1]))
return padded_image, target
class ResizeDebug(object):
def __init__(self, size):
self.size = size
def __call__(self, img, target):
return resize(img, target, self.size)
class RandomCrop(object):
def __init__(self, size):
self.size = size
def __call__(self, img, target):
region = T.RandomCrop.get_params(img, self.size)
return crop(img, target, region)
class RandomSizeCrop(object):
def __init__(self, min_size: int, max_size: int, respect_boxes: bool = False):
# respect_boxes: True to keep all boxes
# False to tolerence box filter
self.min_size = min_size
self.max_size = max_size
self.respect_boxes = respect_boxes
def __call__(self, img: PIL.Image.Image, target: dict):
init_boxes = len(target["boxes"])
max_patience = 10
for i in range(max_patience):
w = random.randint(self.min_size, min(img.width, self.max_size))
h = random.randint(self.min_size, min(img.height, self.max_size))
region = T.RandomCrop.get_params(img, [h, w])
result_img, result_target = crop(img, target, region)
if (
not self.respect_boxes
or len(result_target["boxes"]) == init_boxes
or i == max_patience - 1
):
return result_img, result_target
return result_img, result_target
class CenterCrop(object):
def __init__(self, size):
self.size = size
def __call__(self, img, target):
image_width, image_height = img.size
crop_height, crop_width = self.size
crop_top = int(round((image_height - crop_height) / 2.0))
crop_left = int(round((image_width - crop_width) / 2.0))
return crop(img, target, (crop_top, crop_left, crop_height, crop_width))
class RandomHorizontalFlip(object):
def __init__(self, p=0.5):
self.p = p
def __call__(self, img, target):
if random.random() < self.p:
return hflip(img, target)
return img, target
class RandomResize(object):
def __init__(self, sizes, max_size=None):
assert isinstance(sizes, (list, tuple))
self.sizes = sizes
self.max_size = max_size
def __call__(self, img, target=None):
size = random.choice(self.sizes)
return resize(img, target, size, self.max_size)
class RandomPad(object):
def __init__(self, max_pad):
self.max_pad = max_pad
def __call__(self, img, target):
pad_x = random.randint(0, self.max_pad)
pad_y = random.randint(0, self.max_pad)
return pad(img, target, (pad_x, pad_y))
class RandomSelect(object):
"""
Randomly selects between transforms1 and transforms2,
with probability p for transforms1 and (1 - p) for transforms2
"""
def __init__(self, transforms1, transforms2, p=0.5):
self.transforms1 = transforms1
self.transforms2 = transforms2
self.p = p
def __call__(self, img, target):
if random.random() < self.p:
return self.transforms1(img, target)
return self.transforms2(img, target)
class ToTensor(object):
def __call__(self, img, target):
return F.to_tensor(img), target
class RandomErasing(object):
def __init__(self, *args, **kwargs):
self.eraser = T.RandomErasing(*args, **kwargs)
def __call__(self, img, target):
return self.eraser(img), target
class Normalize(object):
def __init__(self, mean, std):
self.mean = mean
self.std = std
def __call__(self, image, target=None):
image = F.normalize(image, mean=self.mean, std=self.std)
if target is None:
return image, None
target = target.copy()
h, w = image.shape[-2:]
if "boxes" in target:
boxes = target["boxes"]
boxes = box_xyxy_to_cxcywh(boxes)
boxes = boxes / torch.tensor([w, h, w, h], dtype=torch.float32)
target["boxes"] = boxes
return image, target
class Compose(object):
def __init__(self, transforms):
self.transforms = transforms
def __call__(self, image, target):
for t in self.transforms:
image, target = t(image, target)
return image, target
def __repr__(self):
format_string = self.__class__.__name__ + "("
for t in self.transforms:
format_string += "\n"
format_string += " {0}".format(t)
format_string += "\n)"
return format_string