HM-Bench is the first benchmark for MLLMs on hyperspectral images, showing models struggle with complex spatial-spectral reasoning and perform better with visual PCA images than textual reports.
Nvidia nemotron nano v2 vl
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
citation-role summary
baseline 1
citation-polarity summary
years
2026 12roles
baseline 1polarities
baseline 1representative citing papers
Introduces 2M synthetic WATER-S dataset and WATERec model achieving 90.40% accuracy on WordArt-Bench, outperforming prior STR methods and VLMs.
Forensic Knowledge Graphs integrate forensic traces, causal dependencies, and scene links via a new authentication network and Iterative Context Refinement to outperform standard detectors and VLMs on detection, localization, and justification.
PorTEXTO benchmark shows sharp real-world performance drops in pt-PT OCR and finds specialized multilingual data outperforms model size or resolution increases.
RealDocBench supplies 1,356 field-level QA questions over 581 real documents and 1,500 annotated pages, evaluating 18 systems on per-field accuracy, cost, and latency.
MemLens benchmark shows long-context LVLMs lose accuracy with length while memory agents lose visual fidelity, with multi-session reasoning below 30% for most systems and neither approach solving the task alone.
Symmetric Nucleus Subsampling and Expert Embedding Engine reduce modality gaps in multimodal embeddings by over 90% and outperform baselines in data curation for downstream models.
Introduces AMALIA-VL, the first open-source instruction-tuned LVLM for European Portuguese, using a high-resolution vision encoder, pt-PT language model, learned connector, and three-stage training on a custom data mix.
An agentic three-stage video annotation pipeline with an MSTED intermediate and top-down domain adaptation produces CoT training data that lifts Cosmos-Reason2 past Gemini on traffic event reasoning.
LinMU achieves linear-complexity multimodal understanding by swapping self-attention for an M-MATE dual-branch block and distilling from a frozen teacher VLM, matching accuracy with up to 2.7x faster TTFT and 9x higher throughput.
A neuro-symbolic engine generates GeoSym127K, a 127K-question dataset with symbolic ground truths and verified CoT pairs, yielding +22.21% gains on MathVerse Vision-Only after SFT on Qwen3-VL-8B.
citing papers explorer
-
HM-Bench: A Comprehensive Benchmark for Multimodal Large Language Models in Hyperspectral Remote Sensing
HM-Bench is the first benchmark for MLLMs on hyperspectral images, showing models struggle with complex spatial-spectral reasoning and perform better with visual PCA images than textual reports.
-
Advancing WordArt-Oriented Scene Text Recognition: Datasets and Methods
Introduces 2M synthetic WATER-S dataset and WATERec model achieving 90.40% accuracy on WordArt-Bench, outperforming prior STR methods and VLMs.
-
Trustworthy Image Authentication using Forensic Knowledge Graphs
Forensic Knowledge Graphs integrate forensic traces, causal dependencies, and scene links via a new authentication network and Iterative Context Refinement to outperform standard detectors and VLMs on detection, localization, and justification.
-
PorTEXTO: A European Portuguese Benchmark for Visual Text Extraction
PorTEXTO benchmark shows sharp real-world performance drops in pt-PT OCR and finds specialized multilingual data outperforms model size or resolution increases.
-
RealDocBench: A Benchmark for Field-Level QA and Layout Understanding on Real-World Regulated Documents
RealDocBench supplies 1,356 field-level QA questions over 581 real documents and 1,500 annotated pages, evaluating 18 systems on per-field accuracy, cost, and latency.
-
MemLens: Benchmarking Multimodal Long-Term Memory in Large Vision-Language Models
MemLens benchmark shows long-context LVLMs lose accuracy with length while memory agents lose visual fidelity, with multi-session reasoning below 30% for most systems and neither approach solving the task alone.
-
Multimodal Data Curation Through Ranked Retrieval
Symmetric Nucleus Subsampling and Expert Embedding Engine reduce modality gaps in multimodal embeddings by over 90% and outperform baselines in data curation for downstream models.
-
AMALIA-VL: A Native European Portuguese Open-Source Vision and Language Model
Introduces AMALIA-VL, the first open-source instruction-tuned LVLM for European Portuguese, using a high-resolution vision encoder, pt-PT language model, learned connector, and three-stage training on a custom data mix.
-
MAVEN: A Multi-stage Agentic Annotation Pipeline for Video Reasoning Tasks
An agentic three-stage video annotation pipeline with an MSTED intermediate and top-down domain adaptation produces CoT training data that lifts Cosmos-Reason2 past Gemini on traffic event reasoning.
-
LinMU: Multimodal Understanding Made Linear
LinMU achieves linear-complexity multimodal understanding by swapping self-attention for an M-MATE dual-branch block and distilling from a frozen teacher VLM, matching accuracy with up to 2.7x faster TTFT and 9x higher throughput.
-
GeoSym127K: Scalable Symbolically-verifiable Synthesis for Multimodal Geometric Reasoning
A neuro-symbolic engine generates GeoSym127K, a 127K-question dataset with symbolic ground truths and verified CoT pairs, yielding +22.21% gains on MathVerse Vision-Only after SFT on Qwen3-VL-8B.
- DataComp-VLM: Improved Open Datasets for Vision-Language Models