Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

few-shot depth classification

Few-shot depth classification is a computer vision task in which a machine learning model learns to identify and categorize objects or scenes from depth sensor data using only a very small number of labeled training examples per category. Unlike conventional visual classification that relies on standard color images, depth classification processes spatial distance measurements and geometric surface information captured by range sensors or depth cameras. In few-shot scenarios, the model must generalize effectively to target classes despite data scarcity, often by leveraging robust feature representations derived from pre-trained depth encoders or joint multimodal embedding spaces, where a simple classifier can be adapted using just a handful of reference samples.

1 item

ImageBind One Embedding Space to Bind Them All

ImageBind One Embedding Space to Bind Them All

Rohit Girdhar, Alaaeldin El-Nouby, Zhuang Liu, Mannat Singh, Kalyan Vasudev Alwala, Armand Joulin, Ishan Misra

OrganizationsMeta

Why you should read this

Demonstrates that aligning various modalities (audio, depth, thermal) to images automatically aligns them to each other, creating a universal embedding space.

Part of:Must-read papers in multimodal AI

We present ImageBind, an approach to learn a joint embedding across six different modalities - images, text, audio, depth, thermal, and IMU data. We show that all combinations of paired data are not necessary to train such a joint embedding, and only image-paired data is sufficient to bind the modalities together. ImageBind can leverage recent large scale vision-language models, and extends their zero-shot capabilities to new modalities just by using their natural pairing with images. It enables novel emergent applications 'out-of-the-box' including cross-modal retrieval, composing modalities with arithmetic, cross-modal detection and generation. The emergent capabilities improve with the strength of the image encoder and we set a new state-of-the-art on emergent zero-shot recognition tasks across modalities, outperforming specialist supervised models. Finally, we show strong few-shot recognition results outperforming prior work, and that ImageBind serves as a new way to evaluate vision models for visual and non-visual tasks.

Added

2026-01-28