periakiva/finding_berries

PyTorch implementation of the paper "Finding Berries: Segmentation and Counting of Cranberries using Point Supervision and Shape Priors". Peri Akiva, Kristin Dana, Peter Oudemans, Michael Mars. CVPRW2020.

What's up? 👋

Current Topics of Interest

• Self-supervision methods
• Vision transformers
• Instance, panoptic, and semantic segmentation
• Weakly supervised methods
• Generative Adversarial Networks
• Applications in precision agriculture and remote sensing

Recent Publications

• Self-Supervised Object Detection from Egocentric Videos (ICCV 2023)
• Single Stage Weakly Supervised Semantic Segmentation of Complex Scenes (WACV 2023)
• Self-Supervised Material and Texture Representation Learning for Remote Sensing Tasks (CVPR 2022, oral)
• Towards Single Stage Weakly Supervised Semantic Segmentation (Arxiv)
• AI on the Bog: Monitoring and Evaluating Cranberry Crop Risk (WACV 2021)
• H2O-Net: Self-Supervised Flood Segmentation via Adversarial Domain Adaptation and Label Refinement (WACV 2021)
• Finding Berries: Segmentation and Counting of Cranberries using Point Supervision and Shape Priors (CVPRW 2020, oral)
• ViewSynth: Learning Local Features from Depth using View Synthesis (BMVC 2020)

finding_berries's Issues

The data format is correct?

There, we found there is discrepancy between the YAML file versus the provided data,.

In YAML file, the data configure is like this:

data:
image_size: 456x608
local:
type: semantic_seg
model_save_dir: /home/native/projects/data/cranberry/models/
train_dir: /home/native/projects/data/cranberry/points/instance_points/
test_dir: /home/native/projects/data/cranberry/semantic_seg/
val_dir:
docker:
type: semantic_seg
model_save_dir: /app/finding_berries/models/
train_dir: /app/finding_berries/data/instance_points/
test_dir: /app/finding_berries/data/semantic_seg/
val_dir:

Yet, the downloaded data folder structure, we have not found the instance_points folder at all. Please provide insight.

Too many errors

Thank you for your contribution, but there are too many mistakes, I can not run this project, are you still improving this project?