Our Embodied AI (https://embodied-ai.tech/) is also an art piece at #CVPR2026 art gallery!
Watch the backend representation of an AI system that moves a dancer's body with muscle stimulation.
Video: https://youtu.be/iRg6_SKpKrI
(See all art pieces at #CVPR: https://thecvf-art.com/archive.php?year=2026)
RE: https://sigmoid.social/@jaom7/116640689544313929
I will be at #CVPR2026 these days presenting our #TriLite method for weakly-supervised object localization. Get in touch in case you would like to discuss about #WSOL or #AI #Interpretability in general.
Accurate object localization at reduced annotation/computational cost?
#TriLite achieves this by pairing a pre-trained backbone, with our #TriHead component which separates ambiguous, foreground and background features.
The main highlights of #TriLite include:
• State-of-the-art performance
• < 1M trainable parameters
• single-stage training
We will be glad to get in touch at #CVPR2026 to discuss further.
We thank #DEFRA for supporting this research
What a week! 🌟
Our group is celebrating a "triple win" with three papers accepted at #ACL2026, #CVPR2026 Workshop, and #FG2026 !
𝗔𝗖𝗟
𝘗𝘳𝘰𝘛𝘰𝘔: 𝘗𝘳𝘰𝘮𝘰𝘵𝘪𝘯𝘨 𝘗𝘳𝘰𝘴𝘰𝘤𝘪𝘢𝘭 𝘉𝘦𝘩𝘢𝘷𝘪𝘰𝘶𝘳 𝘷𝘪𝘢 𝘛𝘩𝘦𝘰𝘳𝘺 𝘰𝘧 𝘔𝘪𝘯𝘥-𝘐𝘯𝘧𝘰𝘳𝘮𝘦𝘥 𝘍𝘦𝘦𝘥𝘣𝘢𝘤𝘬
𝗖𝗩𝗣𝗥 𝗪𝗼𝗿𝗸𝘀𝗵𝗼𝗽
𝘜𝘗-𝘍𝘢𝘤𝘌: 𝘜𝘴𝘦𝘳-𝘱𝘳𝘦𝘥𝘪𝘤𝘵𝘢𝘣𝘭𝘦 𝘍𝘪𝘯𝘦-𝘨𝘳𝘢𝘪𝘯𝘦𝘥 𝘍𝘢𝘤𝘦 𝘚𝘩𝘢𝘱𝘦 𝘌𝘥𝘪𝘵𝘪𝘯𝘨
𝗙𝗚
𝘋𝘪𝘧𝘧𝘌𝘺𝘦𝘚𝘺𝘯: 𝘜𝘴𝘦𝘳-𝘴𝘱𝘦𝘤𝘪𝘧𝘪𝘤 𝘚𝘶𝘣𝘵𝘭𝘦 𝘌𝘺𝘦 𝘔𝘰𝘷𝘦𝘮𝘦𝘯𝘵 𝘚𝘺𝘯𝘵𝘩𝘦𝘴𝘪𝘴 𝘜𝘴𝘪𝘯𝘨 𝘋𝘪𝘧𝘧𝘶𝘴𝘪𝘰𝘯 𝘔𝘰𝘥𝘦𝘭𝘴
Congratulations to the team!
For preprints and updates, feel free to visit our website: https://www.collaborative-ai.org/
Satellite Earth-observation (EO) time series in the optical and microwave ranges of the electromagnetic spectrum are often irregular due to orbital patterns and cloud obstruction. Compositing addresses these issues but loses information with respect to vegetation phenology, which is critical for many downstream tasks. Instead, we present TESSERA, a pixel-wise foundation model for multi-modal (Sentinel-1/2) EO time series that learns robust, label-efficient embeddings. During model training, TESSERA uses Barlow Twins and sparse random temporal sampling to enforce invariance to the selection of valid observations. We employ two key regularizers: global shuffling to decorrelate spatial neighborhoods and mix-based regulation to improve invariance under extreme sparsity. We find that for diverse classification, segmentation, and regression tasks, TESSERA embeddings deliver state-of-the-art accuracy with high label efficiency, often requiring only a small task head and minimal computation. To democratize access, adhere to FAIR principles, and simplify use, we release global, annual, 10m, pixel-wise int8 embeddings together with open weights/code and lightweight adaptation heads, thus providing practical tooling for large-scale retrieval and inference at planetary scale. The model training/inference code, downstream task code, and pre-generated embeddings can be accessed at https://github.com/ucam-eo
Pedro Lopes
José Oramas
Garrett Lynch IRL
collaborativeai
lalmei
Anil Madhavapeddy