# Align before Adapt: Leveraging Entity-to-Region Alignments for Generalizable Video Action Recognition

arXiv: https://arxiv.org/abs/2311.15619 · HTML: https://mlsanity.com/paper/2311.15619/

## Key facts

- Cited by 12 papers in the map (data through 14 Oct 2026).
- Builds on 62 of its 84 references that are in the map.
- First posted on arXiv 27 Nov 2023.

## Builds on (most cited first, 30 of 62)

- [Attention Is All You Need](https://mlsanity.com/paper/1706.03762/) (2017)
- [Learning Transferable Visual Models From Natural Language Supervision](https://mlsanity.com/paper/2103.00020/) (2021)
- [Hierarchical Text-Conditional Image Generation with CLIP Latents](https://mlsanity.com/paper/2204.06125/) (2022)
- [Neural Discrete Representation Learning](https://mlsanity.com/paper/1711.00937/) (2017)
- [BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation](https://mlsanity.com/paper/2201.12086/) (2022)
- [Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision](https://mlsanity.com/paper/2102.05918/) (2021)
- [Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset](https://mlsanity.com/paper/1705.07750/) (2017)
- [The Kinetics Human Action Video Dataset](https://mlsanity.com/paper/1705.06950/) (2017)
- [SlowFast Networks for Video Recognition](https://mlsanity.com/paper/1812.03982/) (2018)
- [Is Space-Time Attention All You Need for Video Understanding?](https://mlsanity.com/paper/2102.05095/) (2021)
- [ViViT: A Video Vision Transformer](https://mlsanity.com/paper/2103.15691/) (2021)
- [VideoMAE: Masked Autoencoders are Data-Efficient Learners for Self-Supervised Video Pre-Training](https://mlsanity.com/paper/2203.12602/) (2022)
- [CoCa: Contrastive Captioners are Image-Text Foundation Models](https://mlsanity.com/paper/2205.01917/) (2022)
- [Oscar: Object-Semantics Aligned Pre-training for Vision-Language Tasks](https://mlsanity.com/paper/2004.06165/) (2020)
- [Grounded Language-Image Pre-training](https://mlsanity.com/paper/2112.03857/) (2021)
- [The Concrete Distribution: A Continuous Relaxation of Discrete Random Variables](https://mlsanity.com/paper/1611.00712/) (2016)
- [A Closer Look at Spatiotemporal Convolutions for Action Recognition](https://mlsanity.com/paper/1711.11248/) (2017)
- [The "something something" video database for learning and evaluating visual common sense](https://mlsanity.com/paper/1706.04261/) (2017)
- [Video Swin Transformer](https://mlsanity.com/paper/2106.13230/) (2021)
- [Two-Stream Convolutional Networks for Action Recognition in Videos](https://mlsanity.com/paper/1406.2199/) (2014)
- [Token Merging: Your ViT But Faster](https://mlsanity.com/paper/2210.09461/) (2022)
- [VL-BERT: Pre-training of Generic Visual-Linguistic Representations](https://mlsanity.com/paper/1908.08530/) (2019)
- [Open-vocabulary Object Detection via Vision and Language Knowledge Distillation](https://mlsanity.com/paper/2104.13921/) (2021)
- [Temporal Segment Networks: Towards Good Practices for Deep Action Recognition](https://mlsanity.com/paper/1608.00859/) (2016)
- [Multiscale Vision Transformers](https://mlsanity.com/paper/2104.11227/) (2021)
- [TSM: Temporal Shift Module for Efficient Video Understanding](https://mlsanity.com/paper/1811.08383/) (2018)
- [Florence: A New Foundation Model for Computer Vision](https://mlsanity.com/paper/2111.11432/) (2021)
- [X3D: Expanding Architectures for Efficient Video Recognition](https://mlsanity.com/paper/2004.04730/) (2020)
- [FILIP: Fine-grained Interactive Language-Image Pre-Training](https://mlsanity.com/paper/2111.07783/) (2021)
- [ClipCap: CLIP Prefix for Image Captioning](https://mlsanity.com/paper/2111.09734/) (2021)

## Built on by (most cited first, 12 of 12)

- [MM-OR: A Large Multimodal Operating Room Dataset for Semantic Understanding of High-Intensity Surgical Environments](https://mlsanity.com/paper/2503.02579/) (2025)
- [XOV-Action: Towards Generalizable Open-Vocabulary Action Recognition](https://mlsanity.com/paper/2403.01560/) (2024)
- [Learning to Generalize without Bias for Open-Vocabulary Action Recognition](https://mlsanity.com/paper/2502.20158/) (2025)
- [ANNEXE: Unified Analyzing, Answering, and Pixel Grounding for Egocentric Interaction](https://mlsanity.com/paper/2504.01472/) (2025)
- [Storyboard guided Alignment for Fine-grained Video Action Recognition](https://mlsanity.com/paper/2410.14238/) (2024)
- [Towards Scalable Modeling of Compressed Videos for Efficient Action Recognition](https://mlsanity.com/paper/2503.13724/) (2025)
- [VTD-CLIP: Video-to-Text Discretization via Prompting CLIP](https://mlsanity.com/paper/2503.18407/) (2025)
- [DAOS: A Multimodal In-cabin Behavior Monitoring with Driver Action-Object Synergy Dataset](https://mlsanity.com/paper/2601.11990/) (2026)
- [Vision-Based Mistake Analysis in Procedural Activities: A Review of Advances and Challenges](https://mlsanity.com/paper/2510.19292/) (2025)
- [Label-Efficient Skeleton-based Recognition with Stable-Invertible Graph Convolutional Networks](https://mlsanity.com/paper/2511.17345/) (2025)
- [Active Learning for GCN-based Action Recognition](https://mlsanity.com/paper/2511.21625/) (2025)
- [MoVA: Learning Asymmetric Dual Projections for Modular Long Video-Text Alignment](https://mlsanity.com/paper/2607.00858/) (2026)

## Abstract (arXiv)

Large-scale visual-language pre-trained models have achieved significant success in various video tasks. However, most existing methods follow an "adapt then align" paradigm, which adapts pre-trained image encoders to model video-level representations and utilizes one-hot or text embedding of the action labels for supervision. This paradigm overlooks the challenge of mapping from static images to complicated activity concepts. In this paper, we propose a novel "Align before Adapt" (ALT) paradigm. Prior to adapting to video representation learning, we exploit the entity-to-region alignments for each frame. The alignments are fulfilled by matching the region-aware image embeddings to an offline-constructed text corpus. With the aligned entities, we feed their text embeddings to a transformer-based video adapter as the queries, which can help extract the semantics of the most important entities from a video to a vector. This paradigm reuses the visual-language alignment of VLP during adaptation and tries to explain an action by the underlying entities. This helps understand actions by bridging the gap with complex activity semantics, particularly when facing unfamiliar or unseen categories. ALT demonstrates competitive performance while maintaining remarkably low computational costs. In fully supervised experiments, it achieves 88.1% top-1 accuracy on Kinetics-400 with only 4947 GFLOPs. Moreover, ALT outperforms the previous state-of-the-art methods in both zero-shot and few-shot experiments, emphasizing its superior generalizability across various learning scenarios.