# SpectraDINO: Modality-Conditioned Adaptation of RGB Vision Foundation Models Across Infrared Bands

arXiv: https://arxiv.org/abs/2605.02258 · HTML: https://mlsanity.com/paper/2605.02258/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 35 of its 58 references that are in the map.
- First posted on arXiv 4 May 2026.

## Builds on (most cited first, 30 of 35)

- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Segment Anything](https://mlsanity.com/paper/2304.02643/) (2023)
- [DINOv2: Learning Robust Visual Features without Supervision](https://mlsanity.com/paper/2304.07193/) (2023)
- [Representation Learning with Contrastive Predictive Coding](https://mlsanity.com/paper/1807.03748/) (2018)
- [Masked Autoencoders Are Scalable Vision Learners](https://mlsanity.com/paper/2111.06377/) (2021)
- [Emerging Properties in Self-Supervised Vision Transformers](https://mlsanity.com/paper/2104.14294/) (2021)
- [Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation](https://mlsanity.com/paper/1802.02611/) (2018)
- [Sigmoid Loss for Language Image Pre-Training](https://mlsanity.com/paper/2303.15343/) (2023)
- [CBAM: Convolutional Block Attention Module](https://mlsanity.com/paper/1807.06521/) (2018)
- [BEiT: BERT Pre-Training of Image Transformers](https://mlsanity.com/paper/2106.08254/) (2021)
- [Visual Prompt Tuning](https://mlsanity.com/paper/2203.12119/) (2022)
- [Unified Perceptual Parsing for Scene Understanding](https://mlsanity.com/paper/1807.10221/) (2018)
- [ImageBind: One Embedding Space To Bind Them All](https://mlsanity.com/paper/2305.05665/) (2023)
- [Exploring Plain Vision Transformer Backbones for Object Detection](https://mlsanity.com/paper/2203.16527/) (2022)
- [AdaptFormer: Adapting Vision Transformers for Scalable Visual Recognition](https://mlsanity.com/paper/2205.13535/) (2022)
- [Perceiver IO: A General Architecture for Structured Inputs & Outputs](https://mlsanity.com/paper/2107.14795/) (2021)
- [Vision Transformer Adapter for Dense Predictions](https://mlsanity.com/paper/2205.08534/) (2022)
- [SpectralGPT: Spectral Remote Sensing Foundation Model](https://mlsanity.com/paper/2311.07113/) (2023)
- [LLVIP: A Visible-infrared Paired Dataset for Low-light Vision](https://mlsanity.com/paper/2108.10831/) (2021)
- [MMTM: Multimodal Transfer Module for CNN Fusion](https://mlsanity.com/paper/1911.08670/) (2019)
- [Multi-interactive Feature Learning and a Full-time Multi-modality Benchmark for Image Fusion and Segmentation](https://mlsanity.com/paper/2308.02097/) (2023)
- [Delivering Arbitrary-Modal Semantic Segmentation](https://mlsanity.com/paper/2303.01480/) (2023)
- [HyperSIGMA: Hyperspectral Intelligence Comprehension Foundation Model](https://mlsanity.com/paper/2406.11519/) (2024)
- [Multi-modal Gated Mixture of Local-to-Global Experts for Dynamic Image Fusion](https://mlsanity.com/paper/2302.01392/) (2023)
- [Edge-aware Guidance Fusion Network for RGB Thermal Scene Parsing](https://mlsanity.com/paper/2112.05144/) (2021)
- [UniRGB-IR: A Unified Framework for Visible-Infrared Semantic Tasks via Adapter Tuning](https://mlsanity.com/paper/2404.17360/) (2024)
- [DiffV2IR: Visible-to-Infrared Diffusion Model via Vision-Language Understanding](https://mlsanity.com/paper/2503.19012/) (2025)
- [RGB-X Object Detection via Scene-Specific Fusion Modules](https://mlsanity.com/paper/2310.19372/) (2023)
- [InfMAE: A Foundation Model in the Infrared Modality](https://mlsanity.com/paper/2402.00407/) (2024)
- [HAPNet: Toward Superior RGB-Thermal Scene Parsing via Hybrid, Asymmetric, and Progressive Heterogeneous Feature Fusion](https://mlsanity.com/paper/2404.03527/) (2024)

## Abstract (arXiv)

Vision foundation models (VFMs) pretrained on large-scale RGB data provide strong general-purpose representations, yet infrared perception, which is essential for robotics and driving in low light and adverse weather, still relies on backbones built per sensor. Existing transfer methods bridge the visible-infrared gap for only a single band or a single RGB-IR pair, leaving no unified backbone that spans the infrared spectrum. We present SpectraDINO, a modality-conditioned adaptation of an RGB vision foundation model that supports NIR, SWIR, and LWIR with a single shared ViT. All modality-specific behavior is conditioned on the active sensor via lightweight input stems, learned modality embeddings, and per-modality bottleneck adapters, while the transformer weights remain shared. We propose a staged protocol to distill the frozen DINOv2 teacher into a spectral student via cosine distillation, symmetric contrastive alignment, patch-level correspondence, and a queue-based neighborhood-preservation loss, and then selectively fine-tune the deeper blocks to adapt without forgetting RGB priors. Across seven detection and segmentation benchmarks spanning all three infrared bands, this single conditioned backbone matches or surpasses modality-specific methods, improving mIoU over the best published results by 2.6 points on FMB and 1.4 on SemanticRT, and exceeding identically fine-tuned foundation-model baselines by 2.7 mAP on SWIR detection.