-
Hello! I'm so sorry to bother you. I was hoping you could guide me on finetuning a COCO-pretrained Mask R-CNN using HM3D images. I am interested in exploring other semantic segmentation methods and wo…
-
Very impressed with the all new innovative architecture in Detr!
Can you clarify recommendations for training on a custom dataset?
Should we build a model similar to demo and train, or better to use…
-
I have found that the output's format of Mask R-CNN model use predict() from inference_model.py might not match Openvino's converted model output.
https://github.com/opencv/open_model_zoo/blob/mast…
-
### Search before asking
- [X] I have searched the YOLOv5 [issues](https://github.com/ultralytics/yolov5/issues) and [discussions](https://github.com/ultralytics/yolov5/discussions) and found no simi…
-
Hi CV course contributors,
We would love to hear your feedback on the multimodal transfer learning section of the course.
Here's the current general outline with some of the thoughts we've done in t…
-
I keep getting curl: (56) Recv failure: Connection reset by peer.
Changed the memory/CPU in Docker, but still encounter it. Any idea what can cause this?
-
### 🐛 Describe the bug
Python scripts crashes with code 139 when exporting a model using `torch.jit.script`.
```python
import torch
import faulthandler
faulthandler.enable()
neural_networ…
-
# Bug description
I've been trying to run a top down training session for 2 animals with 5 body parts. I keep running into this error as it is about to initialize the training.
`2024-06-12 14:20…
-
作者您好:
我用MNN->tools->MNNPythonOfflineQuant中的python量化工具(python MNN库版本为1.1.6)时,按照提供的代码和说明处理数据和代码。最后nn.compress.train_quant(net, quant_bits=8)一直报错:segmentation fault(core dumped).注释掉该行之后则可以正常运行;请…
-
Hi, I would like to add the segmentation task from PETRv2 to StreamPETR (not to further improve performance, but to implement this task within the StreamPETR framework). Where should I start? Can you …