-
Hi,
We would request if all the backbones for SCNET (https://github.com/open-mmlab/mmdetection/tree/master/configs/scnet) are hosted on openmmlab.com similar to VFNET (for e.g.: https://download.op…
-
## 📚 Documentation
It is possible (although not trivial) to get bounding box augmentations / transforms working with `ObjectDetectionData` (see #414). We should document this somewhere, ideally pro…
-
## 🐛 Bug
One of the dependencies of lightning-flash requires torch version
-
## 🐛 Bug
**Describe the bug**
A clear and concise description of what the bug is.
**To Reproduce**
Steps to reproduce the behavior:
1. Go to '...'
2. Click on '....'
3. Scroll down to '....'
…
-
Update example using icevision 0.8 version
-
## 🐛 Bug
Icevision 0.8.1
```
model_type = models.torchvision.faster_rcnn
backbone = models.torchvision.faster_rcnn.backbones.resnet50
model = model_type.model(backbone=backbone(pretrained=True)…
-
As raised by Farid on Icevision forum, small objects are detected with significantly lower accuracy. The Fire dataset does include many small objects in the form of candle flames, which are not really…
-
Object Detection task currently only seems to support FasterRCNN it would be good to have additional backbones such as RetinaNet or YoloV5 .
-
## 🚀 Feature
IceVision doesn't need all the dependencies needed to train a model. We need a lean script to only install the required dependencies for inference
-
## 🚀 Feature
Progressive resizing is a very effective technique to train model from scratch or using transfer learning. IceVision should offer a good support for that technique