Comments (5)
I also tried with this config:
_base_ = [
'../../_base_/models/retinanet_r50_fpn.py',
'../../_base_/datasets/custom_cocovid_dataset.py',
'../../_base_/default_runtime.py'
]
model = dict(
type='SELSA',
pretrains=None,
detector=dict(
bbox_head=dict(
type='SelsaBBoxHead',
num_shared_fcs=2,
aggregator=dict(
type='SelsaAggregator',
in_channels=1024,
num_attention_blocks=16
),
num_classes=1,
)
)
)
This time Im getting a different error:
Traceback (most recent call last):
File "tools/train.py", line 168, in <module>
main()
File "tools/train.py", line 141, in main
model = build_model(cfg.model)
File "mmtracking\mmtrack\models\builder.py", line 69, in build_model
return build(cfg, MODELS)
File "mmtracking\mmtrack\models\builder.py", line 34, in build
return build_from_cfg(cfg, registry, default_args)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmcv\utils\registry.py", line 171, in build_from_cfg
return obj_cls(**args)
File "mmtracking\mmtrack\models\vid\selsa.py", line 23, in __init__
self.detector = build_detector(detector)
File "mmtracking\mmtrack\models\builder.py", line 60, in build_detector
return build(cfg, DETECTORS)
File "mmtracking\mmtrack\models\builder.py", line 34, in build
return build_from_cfg(cfg, registry, default_args)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmcv\utils\registry.py", line 171, in build_from_cfg
return obj_cls(**args)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmdet\models\detectors\retinanet.py", line 16, in __init__
super(RetinaNet, self).__init__(backbone, neck, bbox_head, train_cfg,
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmdet\models\detectors\single_stage.py", line 30, in __init__
self.bbox_head = build_head(bbox_head)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmdet\models\builder.py", line 59, in build_head
return build(cfg, HEADS)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmdet\models\builder.py", line 34, in build
return build_from_cfg(cfg, registry, default_args)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmcv\utils\registry.py", line 171, in build_from_cfg
return obj_cls(**args)
File "mmtracking\mmtrack\models\roi_heads\bbox_heads\selsa_bbox_head.py", line 19, in __init__
super(SelsaBBoxHead, self).__init__(*args, **kwargs)
File "C:\Users\FCA\Miniconda3\envs\mmtracking\lib\site-packages\mmdet\models\roi_heads\bbox_heads\convfc_bbox_head.py", line 33, in __init__
super(ConvFCBBoxHead, self).__init__(*args, **kwargs)
TypeError: __init__() got an unexpected keyword argument 'stacked_convs'
from mmtracking.
SELSA need aggregate the features of proposals, and RetinaNet doesn't own proposals. Therefore, SELSA cann't be applied into RetinaNet.
from mmtracking.
@GT9505 i see. how to train selsa with proposal generating detectors other than fasterrcnn? config stays the same and i only need to change the base model config?
from mmtracking.
The architecture of second stage in fasterrcnn is RoI Align -> fc1 -> fc2 -> cls/reg
, and selsa is plugged after each fc (RoI Align -> fc1 -> selsa -> fc2 -> selsa -> cls/reg
).
Thus, for the image detectors having the similar architecture with fasterrcnn, you just need to modify the base model config. Otherwise, you need change the code.
from mmtracking.
@GT9505 thanks for the feedback
from mmtracking.
Related Issues (20)
- 您好,您的mmtrack/datasets/pipelines/transforms.py中的SeqExpand应该是写错了,得到的序列中每个图片的增强都不同,在vid算法训练时,时序信息将会丢失。 HOT 2
- mmtracking and BDD100Kdataset
- Multiple static targets in MixFormer
- Poor training results on custom data with ''Temporal ROI Align for Video Object Recognition'' ?
- Maybe a small bug about test progress bar in multi_gpu_test(). HOT 1
- AssertionError: MMCV==2.0.1 is used but incompatible. Please install mmcv>=1.3.17, <2.0.0. HOT 2
- ModuleNotFoundError: No module named 'mmtrack' HOT 1
- mmcv compatibility
- ModuleNotFoundError: No module named 'mmcv._ext' HOT 1
- Failing to build wheels of mmtrack HOT 1
- The MOT tutorial does not output label for detection results
- KeyError: 'categories'
- Using YOLOV8 in MMTracking
- -
- mmdet depends on mmcv>=2.0.0rc4 while mmtrack depends on mmcv <2.0.0 HOT 2
- Compatability Issue between MMCV and MMTracking HOT 1
- in _get_stream if device.type == "cpu": AttributeError: 'int' object has no attribute 'type' HOT 1
- Is this repo dead? HOT 2
- KeyError: "'track_bboxes' not found in the outputs."
- improve compatibility between mmcv and mmdection
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from mmtracking.