Spaces:
No application file
No application file
File size: 3,877 Bytes
907462b | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 | import torch
import datetime
from samhi.inference.evaluation_tools import Evaluation
import argparse
time = datetime.datetime.now().strftime("%Y-%m-%d_%H-%M-%S")
parser = argparse.ArgumentParser()
# General Info
model_types = ["vit_b", "vit_l", "vit_h"]
parser.add_argument("--model_type", type=str, choices=model_types, default="vit_b")
parser.add_argument("--base_model", type=str, default="sam_vit_b_01ec64.pth")
parser.add_argument("--model_dir", type=str, default="/vol/data/models/")
# model_names = ["sam_vit_b_01ec64"]
parser.add_argument("--model_name", type=str, default="model-g9e0e56i:v0")
parser.add_argument("--seed", type=int, default=1)
parser.add_argument("--interactive_or_auto", type=str, choices=["interactive", "auto"], default="interactive")
interactive_models = ["sam", "simpleclick", "ours"]
parser.add_argument("--interactive_model", type=str, default="ours")
auto_models = ["sam", "cellvit", "ours"]
parser.add_argument("--auto_model", type=str, choices=auto_models, default="ours")
parser.add_argument("--max_nr_of_points", type=int, default=6)
# Data Info
datasets = ["BCSS", "CAMELYON", "CellSeg", "CoCaHis", "CoNIC", "CPM", "CRAG", "CryoNuSeg", "GlaS", "ICIA2018",
"Janowczyk", "KPI", "Kumar", "MoNuSAC", "MoNuSeg", "NuClick", "PAIP2023", "PanNuke", "SegPath", "SegPC",
"TIGER", "TNBC", "WSSS4LUAD"]
parser.add_argument("--datasets", nargs='+', choices=datasets, default=["CPM"])
parser.add_argument("--data_dir", type=str, default="/vol/data/histo_datasets/")
parser.add_argument("--cluster", type=str, default="denbi")
parser.add_argument("--image_encoder_size", type=int, default=1024)
parser.add_argument("--batch_size", type=int, default=4)
parser.add_argument("--drop_last", type=bool, default=True)
parser.add_argument("--num_workers", type=int, default=5)
# Prompt Info
parser.add_argument("--prompt_batch_size", type=int, default=10)
parser.add_argument("--nr_of_initial_points", type=int, default=1)
parser.add_argument("--nr_of_initial_positive_points", type=int, default=1)
parser.add_argument("--bbox_shift", type=int, default=10)
# Evaluation Info
parser.add_argument("--nr_of_interactive_points_choices", nargs='+', type=int, default=[0, 1, 2, 3, 4, 5])
parser.add_argument("--prompt_choices", nargs='+', choices=["points", "boxes"], default=["points", "boxes"])
args = parser.parse_args()
inference_config = {
"device": "cuda" if torch.cuda.is_available() else "cpu",
"model_dir": args.model_dir,
"model_name": args.model_name,
}
eval_config = {
"interactive_or_auto": args.interactive_or_auto,
"interactive_model": args.interactive_model,
"auto_model": args.auto_model,
"time": time,
"seed": args.seed,
"batch_size": args.batch_size,
"prompt_batch_size": args.prompt_batch_size,
"max_nr_of_points": args.max_nr_of_points,
"nr_of_interactive_points_choices": args.nr_of_interactive_points_choices,
"prompt_choices": args.prompt_choices,
"dataset_choices": args.datasets,
"mask_threshold": 0.0,
"model_type": args.model_type,
"base_model": args.base_model,
}
## Evaluate on the following datasets: CRAG (glands), MoNuSeg (cells), Cellseg (cells on blood smear)
data_config = {
"data_directory": args.data_dir,
"cluster": args.cluster,
"image_encoder_size": args.image_encoder_size,
"batch_size": args.batch_size,
"drop_last": args.drop_last,
"num_workers": args.num_workers,
}
prompt_config = {
"prompt_batch_size": args.prompt_batch_size,
"nr_of_points": args.nr_of_initial_points,
"nr_of_positive_points": args.nr_of_initial_positive_points,
"bbox_shift": args.bbox_shift,
}
config = {
"inference_config": inference_config,
"eval_config": eval_config,
"data_config": data_config,
"prompt_config": prompt_config,
}
evaluation = Evaluation(config)
evaluation.evaluate()
|