Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2011.14752

A Comprehensive Review on Recent Methods and Challenges of Video Description

Alok Singh·Thoudam Doren Singh·Sivaji Bandyopadhyay
View original source ↗Read source document ↗Export metadata ↓

Abstract

Video description involves the generation of the natural language description of actions, events, and objects in the video. There are various applications of video description by filling the gap between languages and vision for visually impaired people, generating automatic title suggestion based on content, browsing of the video based on the content and video-guided machine translation [86] etc.In the past decade, several works had been done in this field in terms of approaches/methods for video description, evaluation metrics,and datasets. For analyzing the progress in the video description task, a comprehensive survey is needed that covers all the phases of video description approaches with a special focus on recent deep learning approaches. In this work, we report a comprehensive survey on the phases of video description approaches, the dataset for video description, evaluation metrics, open competitions for motivating the research on the video description, open challenges in this field, and future research directions. In this survey, we cover the state-of-the-art approaches proposed for each and every dataset with their pros and cons. For the growth of this research domain,the availability of numerous benchmark dataset is a basic need. Further, we categorize all the dataset into two classes: open domain dataset and domain-specific dataset. From our survey, we observe that the work in this field is in fast-paced development since the task of video description falls in the intersection of computer vision and natural language processing. But still, the work in the video description is far from saturation stage due to various challenges like the redundancy due to similar frames which affect the quality of visual features, the availability of dataset containing more diverse content and availability of an effective evaluation metric.

Explore related subjects

cs.CVcs.CL

Keep this discovery

BibTeXRIS

Alok Singh, Thoudam Doren Singh, Sivaji Bandyopadhyay. 2020-11-30. A Comprehensive Review on Recent Methods and Challenges of Video Description. https://arxiv.org/abs/2011.14752

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

HiPerViT: A Hierarchical Perceiver-Vision Transformer Architecture for Multi-Scale Texture Recognition

Texture recognition remains challenging for modern vision models because discriminative evidence is often carried by higher-order spatial statistics rather than by object shape alone. While Vision Transformers provide strong long-range modeling capacity, their standard object-centric representations do not explicitly expose such statistical structure, which limits texture sensitivity in fine-grained recognition settings. We present HiPerViT, a compact vision-only architecture that injects an explicit second-order statistical prior into a transformer-based recognition pipeline. The method combines global and local image views with a compact bilinear descriptor encoded as a statistical token, and integrates this token with first-order spatial representations through Perceiver-style latent distillation. This design enables direct interaction between spatial tokens and second-order feature co-occurrence statistics, providing the model with explicit access to texture-relevant information without requiring multimodal pretraining or ensemble construction. Across six texture recognition benchmarks, HiPerViT achieves consistent improvements over strong vision-only baselines under the reported evaluation protocols, including gains of +3.05 percentage points on DTD, +10.48 on GTOS-Mobile, and +10.10 on 1200Tex. Beyond benchmark performance, our analyses show that these gains are largely invariant to the backbone depth used to extract second-order statistics and to the ordering of interaction and distillation stages. This pattern suggests that the primary source of improvement is not a specific fusion topology, but the explicit availability of second-order statistical information as a first-class representational signal. These results support explicit statistical tokenization as an effective and robust design principle for texture-centric visual recognition.

cs.CV↗
arXiv2026

CamPilot: A Multi-Agent Cinematic Assistant for Camera-Controlled Movie Generation

The integration of large language models (LLMs) into video generation has enabled rapid text-to-video creation and improved visual quality. However, it still falls short of professional filmmaking, where cinematographic language is less refined than human-crafted camera work and multi-shot continuity remains challenging. To address these limitations, we introduce CamPilot, a multi-agent framework that integrates cinematographic planning and camera-work control to produce more coherent, logically structured, and human-aesthetic movies. CamPilot adopts a GRPO-based learning paradigm to learn camera work planning from 14K real-world professional movies, internalizing motion patterns and composition principles that support reasoning over shooting techniques (e.g., camera angle, motion, and focal behavior) and cross-shot relationships for controllable camera-viewpoint generation. Multiple agents further collaborate and evolve to improve overall output quality. To support this work and further studies in this domain, we establish CamEval, a benchmark for evaluating camera work quality and cinematic engagement. Empirical results show that CamPilot outperforms state-of-the-art text-to-movie generation methods on cinematographic control and quality, highlighting the impact of professional camera design on movie generation.

cs.CV↗
arXiv2026

New Evidence, Same Choice: Testing Physical Experiment Selection in Vision Language Models

A model first sees an image from one physical measurement experiment, such as how far a block coasted, and must answer a question about a new trial, such as whether the block will pass a target after a fixed push. The initial experiment may provide enough information to answer, or the model may need another measurement, such as the object's mass, friction, restitution, or spring stiffness. We study whether vision language models can decide when to answer immediately and, when more evidence is needed, which experiment to perform. Current physical reasoning benchmarks usually evaluate only the final answer, so they do not directly measure this decision-making ability. We introduce a controlled evaluation where each problem provides one measurement image and four possible physical worlds created by combining two possible masses and two possible values of another relevant property. The model must either stop and answer or select the cheapest additional experiment that can resolve the question. We construct matched problem pairs where changing either the observed measurement or the question changes the optimal action. Since all possible worlds and experiment costs are known, we can explicitly determine the optimal choice. Across six open models and 144 physical parameter sets, direct responses repeat the same action for 95.1% to 100% of image pairs even when the correct action changes. Brief reasoning improves action switching, but the best model makes both decisions correctly for only 5.9% of image pairs. Additional analysis reveals failures in measurement interpretation, physical reasoning, and response formatting. By evaluating evidence selection separately from final answers, our benchmark reveals limitations in physical reasoning that conventional answer accuracy can overlook.

cs.CV↗