ShoufaChen / DiffusionDet

[ICCV2023 Oral] PyTorch implementation of DiffusionDet (https://arxiv.org/abs/2211.09788)

ShoufaChen/DiffusionDet Issues

About confidence_threshold
Updated a month ago3
Used Multi gpus but takes same time with 1 gpu training
Updated 2 months ago
evaluator
Updated 2 months ago
AP is nan
Updated 2 months ago
pre-trained models
Updated 2 months ago
coco dataset
Updated 2 months ago
random seed 为什么固定了随机种子和参数后，模型的预测框测试出来都一样
Updated 4 months ago
RuntimeError: nvrtc: error: invalid value for --gpu-architecture (-arch) ON RTX4090
Updated 4 months ago
请问训练时采用的设备是啥呢？
Updated 6 months ago2
复现
Updated 6 months ago3
复现问题
Updated 6 months ago
Too many boxes in
Closed a year ago4
train on my datasets,adjust the value of thread
Closed 6 months ago
使用提供的训练模型做eval只有32mAP
Updated 6 months ago
ap=0
Closed 6 months ago
Questions on box renewal
Updated 7 months ago
Detectron's version of cuda
Updated 7 months ago
不太理解公式2怎么来的
Updated 7 months ago1
why the dataset only split to train and test?
Updated 7 months ago
Prepare datasets
Updated 7 months ago1
Question about `scale` factor
Updated 7 months ago
Training Issue: TypeError: '>' not supported between instances of 'NoneType' and 'int'
Closed 8 months ago1
If freeze the backbone?
Updated 8 months ago
Could you help me understand the diffusion process?
Updated 9 months ago1
When DiffusonDet Model is created in 'train_net.py'?
Closed 9 months ago1
SetCriterionDynamicK and HungarianMatcherDynamicK in loss.py are not well understood
Updated a year ago
How to train on new data after register
Updated a year ago
1
Closed a year ago
Resnet101 worse than Resnet50?
Updated a year ago2
为什么迭代到第三万次的时候会报 CUDA out of memory？
Updated a year ago
config.py不理解
Updated a year ago
请问为什么单卡调试时会提示显存不够？
Closed a year ago1
Understanding proposals
Updated a year ago
Unified input image size
Updated a year ago
我们想测试其他的 encoder对该模型的影响
Closed a year ago
the number of batch size during infering
Closed a year ago1
where is BATCHSIZE? 请问数据的dataload是不是不在主代码库里，放在detectron2里面？
Updated a year ago3
Question about proposals matching
Updated a year ago
An issue when visualizing the model.
Closed a year ago
About the N_train problem mentioned in the paper
Closed a year ago4
An issue about training the Net in VOC
Closed a year ago3
how to run in multi nodes
Updated a year ago1
我的结果比论文结果平均低1.7，请问大家的结果怎么样？
Closed a year ago23
how to focus on a IOU like 0.5 to show the AP of single class?
Updated a year ago
json.decoder.JSONDecodeError: Expecting ',' delimiter: line 1 column 92037121 (char 92037120)
Closed a year ago
UserWarning: FALLBACK path has been taken inside: runCudaFusionGroup. This is an indication that codegen Failed for some reason. To debug try disable codegen fallback path via setting the env variable `export PYTORCH_NVFUSER_DISABLE=fallback` (Triggered internally at ../torch/csrc/jit/codegen/cuda/manager.cpp:331.)
Closed a year ago
"dynamic_k_matching" is better than the origin "linear_sum_assignment" on precision, speed, or both?
Updated a year ago
About the shape of roi_features
Closed a year ago4
About the loss_labels
Updated a year ago
The meaning of "shift" here, and why you need "shift" and "scale"?
Closed a year ago2