TriCLE: Tri-Modal Vision-Language Reasoning for Edge-Deployed Fine-Grained Clustering
Quick Answer
TriCLE is a tri-modal vision-language system designed for edge-deployed aircraft clustering, achieving 88.33% validation accuracy with GSPO strategy.
Quick Take
It processes tri-modal data in 1.48 seconds while fitting an 8GB deployment target, emphasizing the need for further validation with real sensor streams.
Key Points
- TriCLE generates thermal and pseudo-LiDAR views from a single RGB aircraft image.
- Achieved 78.00% accuracy and 0.793 weighted F1 on held-out aircraft test partition.
- Utilizes Qwen3-VL backbone for compact task instruction fusion.
- 4-bit quantization allows the model to fit within an 8GB deployment target.
- Sequence-level GSPO strategy yields the best validation performance.
Paper Resources
📖 Reader Mode
~2 min readAbstract:Edge platforms used for aerial observation must interpret aircraft imagery under limited memory, limited compute, and intermittent connectivity. This setting is difficult for standard RGB-only recognition models and general-purpose vision-language models, especially when calibrated thermal and LiDAR aircraft data are unavailable. We present TriCLE, an application-oriented tri-modal vision-language system for aircraft taxonomic grouping under edge constraints. From a single RGB aircraft image, TriCLE generates a structure-preserving FLIR-style thermal view and a pseudo-LiDAR depth projection, then fuses the aligned views with task instructions in a compact Qwen3-VL backbone. The model is aligned to an expert aircraft taxonomy based on propulsion, airframe family, size, design era, and configuration, so its outputs reflect engineering-relevant similarity rather than only surface appearance. We evaluate supervised fine-tuning, rotation-preserving SFT, and three policy-alignment strategies: GRPO, GSPO, and DAPO. Sequence-level GSPO gives the strongest validation performance, reaching 88.33\% validation accuracy and 0.91 weighted F1 on valid aircraft outputs. On a held-out aircraft test partition, GSPO achieves 78.00\% accuracy and 0.793 weighted F1 while preserving 94.00\% parseable output formatting. After 4-bit quantization and attention-memory optimization, the aligned 4B model fits an 8GB deployment target and processes each tri-modal triplet in 1.48 seconds. These results support TriCLE as a practical prototype for interpretable, edge-feasible aircraft grouping, while emphasizing the need for further validation on real aligned thermal and LiDAR sensor streams.
| Subjects: | Computer Vision and Pattern Recognition (cs.CV) |
| Cite as: | arXiv:2608.04175 [cs.CV] |
| (or arXiv:2608.04175v1 [cs.CV] for this version) | |
| https://doi.org/10.48550/arXiv.2608.04175 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Md. Mahfuzur Rahman [view email]
[v1]
Tue, 4 Aug 2026 19:34:08 UTC (15,177 KB)
— Originally published at arxiv.org
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →ProMoE-FL: Prototype-conditioned Mixture of Experts for Multimodal Federated Learning with Missing Modalities
ProMoE-FL introduces a Prototype-conditioned Mixture-of-Experts framework for multimodal federated learning, effectively addressing missing modalities. It outperforms existing methods on four chest X-ray datasets, demonstrating superior feature synthesis capabilities in both homogeneous and heterogeneous settings.