AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

Falcon: Functional Assembly and Language for Compositional Reasoning in X-ray

Yonathan Michael, Mohamad Alansari, Natnael Takele, Andreas Henschel, Naoufel Werghi

ECCV 2026 · Vision-language and multimodal › Multimodal LLMs · approach: Multimodal LLM benchmarks · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add the structured safety adapter (risk, presence, link heads) to the region-grounded VLM (ablation)Falcon · Falcon-X RFGcIoU: 35.82 -> 50.45Improved
add presence head on top of risk head (ablation)SSA · Falcon-X RFGcIoU: 39.33 -> 45.52Improved
add the link head (ablation)SSA · Falcon-X RFGmIoU: 58.60 -> 69.58Improved
add SSA structured supervision on top of LoRA adaptation (ablation)Falcon · Falcon-X RFGcIoU: 44.63 -> 50.45Improved
swap Vicuna-7B for Qwen3.5-VL backbone (ablation)Falcon · Falcon-X RFGcIoU: 50.45 -> 51.07 (with SSA); 35.82 -> 36.91 (without)No clear effect
use Falcon instead of fine-tuned baselines for referring functional groundingFalcon · Falcon-X RFGcIoU: 20.07 -> 50.45Improved
structured safety bottleneck effect on panoptic segmentationFalcon · Falcon-X PScIoU: 36.73 -> 14.37Hurt
relax the structured bottleneck (ablation)Falcon · Falcon-X PS/RPS vs RFGcIoU: +5.27 PS cIoU, +3.99 RPS cIoU, -7.39 RFG cIoUMixed
increase proposal recall and unfreeze visual encoder (ablation)Falcon · Falcon-X PS/RPScIoU: recall 92.6% -> 97.8%, change under 1 cIoUNo clear effect
replace zero-fill removal with mask-guided inpainting for counterfactuals (ablation)Falcon-X counterfactual images · DINO-FID realism checkDINO-FID: 32.74 -> 6.72Improved
replace Groma with Falcon on missing-component identificationFalcon · Falcon-X MCIaccuracy: 97.1 -> 94.75Hurt
Falcon versus Sting-Bee on cross-dataset referring localizationFalcon · STCray/PIDray/SIXraycIoU: 10.11 -> 16.60 (STCray)Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy