Paper: arxiv.org/abs/2511.18452

by Loïck Chambon (loickch.github.io), @paulcouairon.bsky.social, @eloizablocki.bsky.social, @alexandreboulch.bsky.social, @nicolasthome.bsky.social, @matthieucord.bsky.social

Collab with @mlia-isir.bsky.social

📢 NAF is fully open-source!

The repo contains:
✅ Pretrained model
✅ Example notebooks
✅ Evaluation and training codes

Check it out & ⭐ the repo: github.com/valeoai/NAF

🛠️ Already have a complex, pre-trained pipeline?
If you are using bilinear interpolation anywhere, NAF acts as a strict drop-in replacement.

Just swap it in. No retraining required. It’s literally free points for your metrics.📈

🎯 NAF is versatile!

Not just zero-shot feature upsampling: it shines on image restoration too, delivering sharp, high-quality results across multiple applications. 🖼️

🔬 NAF meets theory.
Under the hood, NAF learns an Inverse Discrete Fourier Transform: revealing a link between feature upsampling, classical filtering, and Fourier theory.

✨ Feature upsampling is no longer a black box

💡 NAF is a super simple, universal architecture that reweights any features using only the high-resolution image:

🧬 Lightweight image encoder (600k params)
🔁 Rotary Position Embeddings (RoPE)
🔍 Cross-Scale Neighborhood Attention

First fully learnable VFM-agnostic reweighting!✅

🔥 NAF sets a new SoTA!
It beats both VFM-specific upsamplers (FeatUp, JAFAR) and VFM-agnostic methods (JBU, AnyUp) across downstream tasks:

- 🥇Semantic Segmentation
- 🥇Depth Estimation
- 🥇Open-Vocabulary
- 🥇Video Propagation, etc.

Even for massive models like: DINOv3-7B !

Improving Multimodal Distillation for 3D Semantic Segmentation under Domain Shift

by: @bjoernmichele.bsky.social @alexandreboulch.bsky.social @gillespuy.bsky.social @tuanhungvu.bsky.social, R. Marlet, @ncourty.bsky.social

📄 bsky.app/profile/bjoe...
Code: ✅

LED: Light Enhanced Depth Estimation at Night

by: S de Moreau, Y. Almehio, @abursuc.bsky.social, H. El-Idrissi, B. Stanciulescu, @fabienmoutarde

tl;dr: a light enhancement method for better depth estimation in low-light conditions

📄 arxiv.org/abs/2409.08031
Code: ✅

LOGen: Toward Lidar Object Generation by Point Diffusion

by: E. Kirby, @mickaelchen.bsky.social, R. Marlet, N. Samet

tl;dr: a diffusion-based method producing lidar point clouds of dataset objects, with an extensive control of the generation

📄 arxiv.org/abs/2412.07385
Code: ✅

Analyzing Fine-tuning Representation Shift for Multimodal LLMs Steering Alignment

tl;dr: a new method for understanding and controlling how MLLMs adapt during fine-tuning

by: P. Khayatan, M. Shukor, J. Parekh, A. Dapogny, @matthieucord.bsky.social
📄: arxiv.org/abs/2501.03012

FLOSS: Free Lunch in Open-vocabulary Semantic Segmentation

tl;dr: a simple trick to boost open-vocabulary semantic segmentation by identifying class expert prompt templates

by: Y. Benigmim, M. Fahes, @tuanhungvu.bsky.social, @abursuc.bsky.social, R. de Charette.
📄: arxiv.org/abs/2504.10487

MoSiC: Optimal-Transport Motion Trajectories for Dense Self-Supervised Learning

tl;dr: a self-supervised learning of temporally consistent representations from video w/ motion cues

by: M. Salehi, S. Venkataramanan, I. Simion, E. Gavves, @cgmsnoek.bsky.social, Y. Asano
📄: arxiv.org/abs/2506.08694

GaussRender: Learning 3D Occupancy with Gaussian Rendering

tl;dr: a module for 3D occupancy learning that enforces 2D-3D consistency through differentiable Gaussian rendering

by: L. Chambon, @eloizablocki.bsky.social, @alexandreboulch.bsky.social, M. Chen, M. Cord
📄: arxiv.org/abs/2502.05040

DIP: Unsupervised Dense In-Context Post-training of Visual Representations

@ssirko.bsky.social, @vobeckya.bsky.social, @abursuc.bsky.social , N. Thome, @spyrosgidaris.bsky.social
📄: arxiv.org/abs/2506.18463

bsky.app/profile/ssir...