Strengths2
- Modular implementation of complex model components (PatchCvT, DETR variant, memory bank) from primitives.
- Complete training and evaluation pipeline with unit tests for core components.
Architecture Reimplementation · Python
patch-DEtection-TRansformer
What it is
The repository implements PatchDETR, a memory-conditioned vision transformer for object detection, combining PatchCvT and DETR architectures with a memory module, along with training and evaluation code.
Merit describes the code at this commit. Later pushes count once the owner rescans. How merit is scored
Where it counts
- Machine Learning80%
- Computer Vision80%
- MLOps30%
- Data Science20%
How much of the project is about each field, as scoring counts it. It ranks in the field it is most about.
What the judge found
Read from the project's code at its scored commit.
Limitations
The owner keeps this project's limitations private.
Around it on the Machine Learning board
Whole board- 63, shared rankmodern-training-pipelineAI Developer Tool · Python56.7ImplementationMohamed KhayatMachine Learning
- 65, shared rankautomatos-academyEducational Implementation · JavaScript56.2ImplementationAutomatos AIMachine Learning
- 65, shared rankpcap-ml-traffic-classifierML Application · Python56.2ImplementationPop AlexandruMachine Learning
- 65, shared rankpatch-DEtection-TRansformerArchitecture Reimplementation · Python56.2ImplementationSuceveanu Mircea Dan SanduMachine Learning
- 68bf16-update-realization-in-transformersTraining Experiment · Python55.8ImplementationRickMachine Learning
More from Suceveanu Mircea Dan Sandu
- Genetic-Algorithm-Based-Recovery-to-a-Nominal-Planner-PathResearch Reproduction · Advanced61.5Machine Learning
- patch-convolutional-vision-transformerML Application · Advanced60.3Machine Learning
- Advanced-Predictive-Railway-Adhesion-Management-SystemArchitecture Reimplementation · Implementation54.3Machine Learning
- AgroADASML Application · Implementation52.6Machine Learning