GSoC 2025 Project Ideas

Please ask questions through issues on the respective project's repo.

Tags available @henrykironde, @bw4sz, @ethanwhite,

• Preferred names (Henry, Ben, Ethan)
• Preferred_greeting (Hi|Hello|Dear|Thanks|Thank you [First_name])

The code of conduct should be your first read.

---

Proposal 1: Efficient Detection of Unique Images from Overlapping Images

Rationale:

Develop a workflow to compute unique image detections from overlapping images using either the weecology/DoubleCounting repository or the open-forest-observatory/geograypher repository. This project will focus on implementing an efficient algorithm for removing double counting among overlapping images.

Approach:

• Choose a suitable repository either weecology/DoubleCounting or open-forest-observatory/geographer for implementing the unique image detections workflow.
• Develop an efficient algorithm for removing double counting among overlapping images.
• Evaluate the performance of the workflow on various datasets.

Expected Outcomes:

• A workflow for computing unique image detections from overlapping images.
• Documentation on using the workflow.

Source Code: DeepForest

Degree of Difficulty:

• Intermediate, long (350 hours)

Skills:

• Deep learning
• Git/GitHub
• Machine learning
• Software testing
• Python and Python package deployment

Mentors:

• @bw4sz
• @henrysenyondo
• @ethanwhite

---

Proposal 2: Integrating PyTorch Wildfire Weights into DeepForest for Airborne Imagery Analysis

Rationale:

Integrate the pre-trained model weights from PyTorch Wildfire into DeepForest, enabling the use of PyTorch Wildfire's models within the DeepForest framework. This project will focus on developing a robust and efficient interface between the two packages, allowing users to leverage the strengths of both frameworks for airborne imagery analysis.

Approach:

• Develop a Python package that allows users to load PyTorch Wildfire model weights into DeepForest.
• Implement a converter to transform PyTorch Wildfire model weights into DeepForest's model format.
• Ensure model compatibility and accuracy when using PyTorch Wildfire models within DeepForest.
• Optimize the integration for performance and usability.

Expected Outcomes:

• A Python interface for integrating PyTorch Wildfire model weights with DeepForest.
• Documentation on using PyTorch Wildfire models within DeepForest, including examples and best practices.
• A tutorial or example code snippet demonstrating the integration and usage of PyTorch Wildfire models within DeepForest.

Source Code: DeepForest

Degree of Difficulty:

• Intermediate, long (350 hours)

Skills:

• Deep learning
• Git/GitHub
• PyTorch
• Model conversion techniques
• Python and Python package deployment

Mentors:

• @bw4sz
• @henrysenyondo
• @ethanwhite

---

Proposal 3: Developing an Active Learning Module for DeepForest

Rationale:

Implement an active learning module for DeepForest, allowing users to select new images for model training based on current model scores. This project will focus on integrating the BOEM repository's active learning code into DeepForest, enabling more efficient model training and improved accuracy.

Approach:

• Integrate the BOEM repository's active learning code into DeepForest.
• Develop a user-friendly interface for selecting new images based on model scores.
• Evaluate the effectiveness of the active learning module in improving model accuracy.

Expected Outcomes:

• An active learning module for DeepForest using BOEM.
• Documentation on using the active learning module.

Source Code: DeepForest

Degree of Difficulty:

• Intermediate, long (350 hours)

Skills:

• Deep learning
• Git/GitHub
• Active learning
• Python and Python package deployment

Mentors:

• @bw4sz
• @henrysenyondo
• @ethanwhite

---

Proposal 4: DeepForest Vision Agent connection with LandingAI

https://github.com/landing-ai/vision-agent?tab=readme-ov-file

Rationale:

Text-based queries of images for labeling and organization.

Approach:

• Create configuration for DeepForest users to register LLM keys
• Object detection and segmentation workflows
• Develop a user-friendly interface for selecting new images based on agent responses
• Evaluate the effectiveness of the active learning module in improving model accuracy.

Expected Outcomes:

• An agent-interaction module for DeepForest using VisionAgent.
• Documentation on using the agent module.

Source Code: DeepForest

Degree of Difficulty:

• Intermediate, long (350 hours)

Skills:

• Deep learning
• Git/GitHub
• Active learning
• Python and Python package deployment

Mentors:

• @bw4sz
• @henrysenyondo
• @ethanwhite

---