# Active Multimodal Distillation for Few-shot Action Recognition

arXiv: https://arxiv.org/abs/2506.13322 · HTML: https://mlsanity.com/paper/2506.13322/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 26 of its 41 references that are in the map.
- First posted on arXiv 16 Jun 2025.

## Builds on (most cited first, 26 of 26)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [Relational Knowledge Distillation](https://mlsanity.com/paper/1904.05068/) (2019)
- [Multimodal Few-Shot Learning with Frozen Language Models](https://mlsanity.com/paper/2106.13884/) (2021)
- [Similarity-Preserving Knowledge Distillation](https://mlsanity.com/paper/1907.09682/) (2019)
- [Few-Shot Object Detection with Attention-RPN and Multi-Relation Detector](https://mlsanity.com/paper/1908.01998/) (2019)
- [Cross Modal Distillation for Supervision Transfer](https://mlsanity.com/paper/1507.00448/) (2015)
- [AIM: Adapting Image Models for Efficient Video Action Recognition](https://mlsanity.com/paper/2302.03024/) (2023)
- [Few-Shot Video Classification via Temporal Alignment](https://mlsanity.com/paper/1906.11415/) (2019)
- [Improved Few-Shot Visual Classification](https://mlsanity.com/paper/1912.03432/) (2019)
- [CRNet: Cross-Reference Networks for Few-Shot Segmentation](https://mlsanity.com/paper/2003.10658/) (2020)
- [Temporal-Relational CrossTransformers for Few-Shot Action Recognition](https://mlsanity.com/paper/2101.06184/) (2021)
- [Modality Distillation with Multiple Stream Networks for Action Recognition](https://mlsanity.com/paper/1806.07110/) (2018)
- [Action Recognition with Trajectory-Pooled Deep-Convolutional Descriptors](https://mlsanity.com/paper/1505.04868/) (2015)
- [Spatio-temporal Relation Modeling for Few-shot Action Recognition](https://mlsanity.com/paper/2112.05132/) (2021)
- [Depth Guided Adaptive Meta-Fusion Network for Few-shot Video Recognition](https://mlsanity.com/paper/2010.09982/) (2020)
- [Hybrid Relation Guided Set Matching for Few-shot Action Recognition](https://mlsanity.com/paper/2204.13423/) (2022)
- [Reinforcement Learning through Active Inference](https://mlsanity.com/paper/2002.12636/) (2020)
- [ProtoGAN: Towards Few Shot Learning for Action Recognition](https://mlsanity.com/paper/1909.07945/) (2019)
- [Learning an Augmented RGB Representation with Cross-Modal Knowledge Distillation for Action Detection](https://mlsanity.com/paper/2108.03619/) (2021)
- [Enriching Multimodal Sentiment Analysis through Textual Emotional Descriptions of Visual-Audio Content](https://mlsanity.com/paper/2412.10460/) (2024)
- [Two-stream Collaborative Learning with Spatial-Temporal Attention for Video Classification](https://mlsanity.com/paper/1711.03273/) (2017)
- [Premier-TACO is a Few-Shot Policy Learner: Pretraining Multitask Representation via Temporal Action-Driven Contrastive Loss](https://mlsanity.com/paper/2402.06187/) (2024)
- [Self Paced Adversarial Training for Multimodal Few-shot Learning](https://mlsanity.com/paper/1811.09192/) (2018)
- [A Review of Human Emotion Synthesis Based on Generative Technology](https://mlsanity.com/paper/2412.07116/) (2024)
- [Adaptive Stream Processing on Edge Devices through Active Inference](https://mlsanity.com/paper/2409.17937/) (2024)

## Abstract (arXiv)

Owing to its rapid progress and broad application prospects, few-shot action recognition has attracted considerable interest. However, current methods are predominantly based on limited single-modal data, which does not fully exploit the potential of multimodal information. This paper presents a novel framework that actively identifies reliable modalities for each sample using task-specific contextual cues, thus significantly improving recognition performance. Our framework integrates an Active Sample Inference (ASI) module, which utilizes active inference to predict reliable modalities based on posterior distributions and subsequently organizes them accordingly. Unlike reinforcement learning, active inference replaces rewards with evidence-based preferences, making more stable predictions. Additionally, we introduce an active mutual distillation module that enhances the representation learning of less reliable modalities by transferring knowledge from more reliable ones. Adaptive multimodal inference is employed during the meta-test to assign higher weights to reliable modalities. Extensive experiments across multiple benchmarks demonstrate that our method significantly outperforms existing approaches.