LEMMA: Laplacian pyramids for Efficient Marine SeMAntic Segmentation

Dimensions overall score 8.0

GitHub Code Pulse

No public code linked for this paper yet.

• LEMMA reduces trainable parameters by up to 71x compared to existing models.result 90%
• LEMMA reduces GFLOPs by up to 88.5% compared to existing models.result 90%
• LEMMA reduces inference time by up to 84.65% compared to existing models.result 90%
• LEMMA achieves 93.42% IoU on the Oil Spill dataset.result 90%
• LEMMA achieves 98.97% mIoU on the Mastr1325 dataset.result 90%
• LEMMA uses Laplacian Pyramids to enhance edge recognition for semantic segmentation in marine environments.method 85%
• LEMMA integrates edge information early in the feature extraction process to eliminate computationally expensive feature map computations in deeper layers.technical 85%
• The model may face challenges with data variability and edge cases in natural settings, potentially affecting generalization across unseen data.limitation 80%

Competitive landscape

Competitor map is still being generated for this paper. Enable generation or check back soon.

Startup potential card