Long-term Pre-training for Temporal Action Detection with Transformers

• URL: http://arxiv.org/abs/2408.13152v2
• Date: Mon, 9 Sep 2024 16:16:19 GMT
• Title: Long-term Pre-training for Temporal Action Detection with Transformers
• Authors: Jihwan Kim, Miso Lee, Jae-Pil Heo,
• Abstract summary: Temporal action detection (TAD) is challenging, yet fundamental for real-world video applications. In this paper, we identify two crucial problems from data scarcity: attention collapse and imbalanced performance. We propose a new pre-training strategy, Long-Term Pre-training, tailored for transformers.
• Score: 21.164101507575186
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Temporal action detection (TAD) is challenging, yet fundamental for real-world video applications. Recently, DETR-based models for TAD have been prevailing thanks to their unique benefits. However, transformers demand a huge dataset, and unfortunately data scarcity in TAD causes a severe degeneration. In this paper, we identify two crucial problems from data scarcity: attention collapse and imbalanced performance. To this end, we propose a new pre-training strategy, Long-Term Pre-training (LTP), tailored for transformers. LTP has two main components: 1) class-wise synthesis, 2) long-term pretext tasks. Firstly, we synthesize long-form video features by merging video snippets of a target class and non-target classes. They are analogous to untrimmed data used in TAD, despite being created from trimmed data. In addition, we devise two types of long-term pretext tasks to learn long-term dependency. They impose long-term conditions such as finding second-to-fourth or short-duration actions. Our extensive experiments show state-of-the-art performances in DETR-based methods on ActivityNet-v1.3 and THUMOS14 by a large margin. Moreover, we demonstrate that LTP significantly relieves the data scarcity issues in TAD.

Related papers

• Harnessing Temporal Causality for Advanced Temporal Action Detection [53.654457142657236]
  We introduce CausalTAD, which combines causal attention and causal Mamba to achieve state-of-the-art performance on benchmarks. We ranked 1st in the Action Recognition, Action Detection, and Audio-Based Interaction Detection tracks at the EPIC-Kitchens Challenge 2024, and 1st in the Moment Queries track at the Ego4D Challenge 2024.
  arXiv  Detail & Related papers  (2024-07-25T06:03:02Z)
• Transform-Equivariant Consistency Learning for Temporal Sentence Grounding [66.10949751429781]
  We introduce a novel Equivariant Consistency Regulation Learning framework to learn more discriminative representations for each video. Our motivation comes from that the temporal boundary of the query-guided activity should be consistently predicted. In particular, we devise a self-supervised consistency loss module to enhance the completeness and smoothness of the augmented video.
  arXiv  Detail & Related papers  (2023-05-06T19:29:28Z)
• DuETT: Dual Event Time Transformer for Electronic Health Records [14.520791492631114]
  We introduce the DuETT architecture, an extension of Transformers designed to attend over both time and event type dimensions. DuETT uses an aggregated input where sparse time series are transformed into a regular sequence with fixed length. Our model outperforms state-of-the-art deep learning models on multiple downstream tasks from the MIMIC-IV and PhysioNet-2012 EHR datasets.
  arXiv  Detail & Related papers  (2023-04-25T17:47:48Z)
• Improved Test-Time Adaptation for Domain Generalization [48.239665441875374]
  Test-time training (TTT) adapts the learned model with test data. This work addresses two main factors: selecting an appropriate auxiliary TTT task for updating and identifying reliable parameters to update during the test phase. We introduce additional adaptive parameters for the trained model, and we suggest only updating the adaptive parameters during the test phase.
  arXiv  Detail & Related papers  (2023-04-10T10:12:38Z)
• Rethink Long-tailed Recognition with Vision Transformers [18.73285611631722]
  Vision Transformers (ViT) are hard to train with long-tailed data. ViT learns generalized features in an unsupervised manner. Predictive Distribution (PDC) is a novel metric for Long-Tailed Recognition.
  arXiv  Detail & Related papers  (2023-02-28T03:36:48Z)
• What Makes Data-to-Text Generation Hard for Pretrained Language Models? [17.07349898176898]
  Expressing natural language descriptions of structured facts or relations -- data-to-text generation (D2T) -- increases the accessibility of structured knowledge repositories. Previous work shows that pre-trained language models(PLMs) perform remarkably well on this task after fine-tuning on a significant amount of task-specific training data. We conduct an empirical study of both fine-tuned and auto-regressive PLMs on the DART multi-domain D2T dataset.
  arXiv  Detail & Related papers  (2022-05-23T17:58:39Z)
• BasicTAD: an Astounding RGB-Only Baseline for Temporal Action Detection [46.37418710853632]
  We study a simple, straightforward, yet must-known baseline given the current status of complex design and low detection efficiency in TAD. We extensively investigate the existing techniques in each component for this baseline, and more importantly, perform end-to-end training over the entire pipeline. This simple BasicTAD yields an astounding and real-time RGB-Only baseline very close to the state-of-the-art methods with two-stream inputs.
  arXiv  Detail & Related papers  (2022-05-05T15:42:56Z)
• End-to-end Temporal Action Detection with Transformer [86.80289146697788]
  Temporal action detection (TAD) aims to determine the semantic label and the boundaries of every action instance in an untrimmed video. Here, we construct an end-to-end framework for TAD upon Transformer, termed textitTadTR. Our method achieves state-of-the-art performance on HACS Segments and THUMOS14 and competitive performance on ActivityNet-1.3.
  arXiv  Detail & Related papers  (2021-06-18T17:58:34Z)
• Augmented Transformer with Adaptive Graph for Temporal Action Proposal Generation [79.98992138865042]
  We present an augmented transformer with adaptive graph network (ATAG) to exploit both long-range and local temporal contexts for TAPG. Specifically, we enhance the vanilla transformer by equipping a snippet actionness loss and a front block, dubbed augmented transformer. An adaptive graph convolutional network (GCN) is proposed to build local temporal context by mining the position information and difference between adjacent features.
  arXiv  Detail & Related papers  (2021-03-30T02:01:03Z)
• Convolutional Tensor-Train LSTM for Spatio-temporal Learning [116.24172387469994]
  We propose a higher-order LSTM model that can efficiently learn long-term correlations in the video sequence. This is accomplished through a novel tensor train module that performs prediction by combining convolutional features across time. Our results achieve state-of-the-art performance-art in a wide range of applications and datasets.
  arXiv  Detail & Related papers  (2020-02-21T05:00:01Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.