We propose to improve the time-sensitive video understanding (TSV) capability of video large language models (Video-LLMs) with grounded objects (GO). We hypothesize that TSV tasks can benefit from GO within frames, which is supported by our preliminary experiments on LITA, a state-of-the-art Video-LLM for reasoning temporal localization. While augmenting prompts with textual descriptions of these object annotations improves the performance of LITA, it also introduces extra token length and susceptibility to the noise in object-level information. To address this, we propose GO-Tokenizer, a lightweight add-on module for Video-LLMs leveraging off-the-shelf object detectors to encode compact object information on the fly. Experimental results demonstrate that pretraining with GO-Tokenizer outperforms the vanilla Video-LLM and its counterpart, utilizing textual descriptions of objects in the prompt. The gain generalizes across different models, datasets, and video understanding tasks, such as reasoning temporal localization and dense captioning.

As Multimodal Large Language Models (MLLMs) are increasingly deployed in decision-critical pipelines such as robotics, embodied AI, and safety monitoring, the opacity of their spatial judgments limits operator trust and auditability. MLLMs demonstrate strong reasoning but often struggle with fine-grained spatial understanding and object hallucination. Prior work, ByDeWay, introduced Layered-Depth-Based Prompting (LDP), a training-free framework that mitigates hallucinations by structuring prompts using monocular depth estimation. However, coarse depth layering falls short in resolving object-to-object spatial relationships within the same geometric plane, such as projective ("left of", "above") and topological ("inside", "touching") relations. We propose ByDeWay-V2, which integrates explicit spatial relational context alongside depth cues, expressed as human-readable predicates that serve as auditable evidence for downstream decision support. Using an open-vocabulary object detector (YOLO-World-L), our framework computes pairwise geometric relations between detected objects and injects them as structured spatial predicates into the MLLM prompt, bridging 3D scene depth and 2D spatial semantics without any training. We evaluate ByDeWay-V2 on the Visual Spatial Reasoning (VSR) and BLINK benchmarks across multiple MLLMs, with hallucination grounding assessed via POPE. On the BLINK spatial subset, ByDeWay-V2 achieves a 46 percent relative F1 improvement over LDP for Qwen2.5-VL, and recovers BLIP-Base's spatial reasoning on VSR from near-random performance to a competitive F1 of 0.53. Our lightest configuration operates under a strict 40-token context budget on CPU, showing the framework's suitability for resource-constrained, real-time decision-support settings.

Understanding instrument-tissue interactions is essential for context-aware surgical AI and autonomous robotic surgery. Pretrained vision-language models (VLMs) and vision encoders offer an alternative to conventional interaction classifiers by transferring broad visual and semantic knowledge. However, adapting them to fine-grained surgical interactions remains challenging: (1) freezing the vision encoder depends entirely on pretrained representations that may retain noise and provide weak spatial localization, while (2) full fine-tuning can improve global semantic alignment without ensuring that the encoder learns meaningful features in the correct action region. We address these limitations by introducing LAViFiT, an end-to-end latent-action-guided framework for vision-language fine-tuning. An inverse dynamics model captures the visual changes induced by each action, while a forward world model drives the encoder to represent action-relevant regions. A patch-level SIG Regularizer further prevents local feature collapse without additional supervision, such as bounding boxes or pseudo-labels. Experiments across multiple encoders and datasets improve recognition and image-text alignment, while representation analyses show stronger grounding over the complete instrument-tissue interaction region and more spatially coherent features.

We present Search2Motion, a training-free framework for object-level motion editing in image-to-video generation. Unlike prior methods requiring trajectories, bounding boxes, masks, or motion fields, Search2Motion adopts target-frame-based control, leveraging first-last-frame motion priors to realize object relocation while preserving scene stability without fine-tuning. Reliable target-frame construction is achieved through semantic-guided object insertion and robust background inpainting. We further show that early-step self-attention maps predict object and camera dynamics, offering interpretable user feedback and motivating ACE-Seed (Attention Consensus for Early-step Seed selection), a lightweight search strategy that improves motion fidelity without look-ahead sampling or external evaluators. Noting that existing benchmarks conflate object and camera motion, we introduce S2M-DAVIS and S2M-OMB for stable-camera, object-only evaluation, alongside FLF2V-obj metrics that isolate object artifacts without requiring ground-truth trajectories. Search2Motion consistently outperforms baselines on FLF2V-obj and VBench.

We propose VC-Inspector, a lightweight, open-source large multimodal model (LMM) for reference-free evaluation of video captions, with a focus on factual accuracy. Unlike existing metrics that suffer from limited context handling, weak factuality assessment, or reliance on proprietary services, VC-Inspector offers a reproducible, fact-aware alternative that aligns closely with human judgments. To enable robust training and interpretable evaluation, we introduce a systematic approach for generating captions with controllable errors, paired with graded quality scores and explanatory annotations. Experiments show that VC-Inspector achieves state-of-the-art correlation with human judgments, generalizing across diverse domains (e.g., VATEX-Eval, Flickr8K-Expert, and Flickr8K-CF benchmarks) and revealing the potential for caption improvement.

The risk of misusing text-to-image generative models for malicious uses, especially due to the open-source development of such models, has become a serious concern. As a risk mitigation strategy, attributing generative models with neural fingerprinting is emerging as a popular technique. There has been a plethora of recent work that aim for addressing neural fingerprinting. A trade-off between the attribution accuracy and generation quality of such models has been studied extensively. None of the existing methods yet achieved 100

Video understanding typically requires fine-tuning the large backbone when adapting to new domains. In this paper, we leverage the egocentric video foundation models (EgoVFMs) based on video-language pre-training and propose a parameter-efficient adaptation for egocentric video tasks, namely Ego-VPA. It employs a local sparse approximation for each video frame/text feature using the basis prompts, and the selected basis prompts are used to synthesize video/text prompts. Since the basis prompts are shared across frames and modalities, it models context fusion and cross-modal transfer in an efficient fashion. Experiments show that EgoVPA excels in lightweight adaptation (with only 0.84% learnable parameters), largely improving over baselines and reaching the performance of full fine-tuning.

In this paper, we introduce VideoNarrator, a novel training-free pipeline designed to generate dense video captions that offer a structured snapshot of video content. These captions offer detailed narrations with precise timestamps, capturing the nuances present in each segment of the video. Despite advancements in multimodal large language models (MLLMs) for video comprehension, these models often struggle with temporally aligned narrations and tend to hallucinate, particularly in unfamiliar scenarios. VideoNarrator addresses these challenges by leveraging a flexible pipeline where off-the-shelf MLLMs and visual-language models (VLMs) can function as caption generators, context providers, or caption verifiers. Our experimental results demonstrate that the synergistic interaction of these components significantly enhances the quality and accuracy of video narrations, effectively reducing hallucinations and improving temporal alignment. This structured approach not only enhances video understanding but also facilitates downstream tasks such as video summarization and video question answering, and can be potentially extended for advertising and marketing applications.

We pose keystep recognition as a node classification task, and propose a flexible graph-learning framework for fine-grained keystep recognition that is able to effectively leverage long-term dependencies in egocentric videos. Our approach, termed GLEVR, consists of constructing a graph where each video clip of the egocentric video corresponds to a node. The constructed graphs are sparse and computationally efficient, outperforming existing larger models substantially. We further leverage alignment between egocentric and exocentric videos during training for improved inference on egocentric videos, as well as adding automatic captioning as an additional modality. We consider each clip of each exocentric video (if available) or video captions as additional nodes during training. We examine several strategies to define connections across these nodes. We perform extensive experiments on the Ego-Exo4D dataset and show that our proposed flexible graph-based framework notably outperforms existing methods.

Egocentric videos capture scenes from a wearer's viewpoint, resulting in dynamic backgrounds, frequent motion, and occlusions, posing challenges to accurate keystep recognition. We propose a flexible graph-learning framework for fine-grained keystep recognition that is able to effectively leverage long-term dependencies in egocentric videos, and leverage alignment between egocentric and exocentric videos during training for improved inference on egocentric videos. Our approach consists of constructing a graph where each video clip of the egocentric video corresponds to a node. During training, we consider each clip of each exocentric video (if available) as additional nodes. We examine several strategies to define connections across these nodes and pose keystep recognition as a node classification task on the constructed graphs. We perform extensive experiments on the Ego-Exo4D dataset and show that our proposed flexible graph-based framework notably outperforms existing methods by more than 12 points in accuracy. Furthermore, the constructed graphs are sparse and compute efficient. We also present a study examining on harnessing several multimodal features, including narrations, depth, and object class labels, on a heterogeneous graph and discuss their corresponding contribution to the keystep recognition performance.

Egocentric videos capture scenes from a wearer's viewpoint, resulting in dynamic backgrounds, frequent motion, and occlusions, posing challenges to accurate keystep recognition. We pose keystep recognition as a node classification task, and propose a flexible graph-learning framework for fine-grained keystep recognition that is able to effectively leverage long-term dependencies in egocentric videos. Our approach, termed GLEVR, consists of constructing a graph where each video clip of the egocentric video corresponds to a node. The constructed graphs are sparse and compute efficient, outperforming existing larger models substantially. We further leverage alignment between egocentric and exocentric videos during training for improved inference on egocentric videos, as well as adding automatic captioning as additional modality. We consider each clip of each exocentric video (if available) or video captions as additional nodes during training. We examine several strategies to define connections across these nodes. We perform extensive experiments on the Ego-Exo4D dataset and show that our proposed flexible graph-based framework notably outperforms existing methods by more than 12 points in accuracy on the test server. Our code is publicly available at https://github.com/geohai/graphs-for-keystep-recognition.

We introduce a training-free framework, ByDeWay, a training-free method to boost the performances of Multimodal Large Language Models. Specifically, ByDeWay leverages a novel prompting strategy, Layered-Depth-Based Prompting (LDP), that enhances the spatial reasoning and grounding capabilities of Multimodal Large Language Models (MLLMs). Our key insight is to inject structured spatial context derived from monocular depth estimation into the input prompts-without modifying any model parameters. By segmenting scenes into closest, mid-range, and farthest depth layers and generating region-specific captions using a grounded vision-language model, we produce explicit depth-aware textual descriptions. These descriptions are concatenated with image-question prompts to guide the model toward spatially grounded and hallucination-resistant outputs. Our method is lightweight, modular, and compatible with any black-box MLLM. Evaluations on hallucination-sensitive (POPE) and reasoning-intensive (GQA) tasks show consistent improvements across multiple MLLMs, demonstrating the effectiveness of depth-aware prompting in a zero-training setup. The source code is available at https://github.com/Rajarshi12321/ByDeWay

The design of activation functions remains a pivotal component in optimizing deep neural networks. While prevailing choices like Swish and GELU demonstrate considerable efficacy, they often exhibit domain-specific optima. This work introduces SG-Blend, a novel activation function that blends our proposed SSwish, a first-order symmetric variant of Swish and the established GELU through dynamic interpolation. By adaptively blending these constituent functions via learnable parameters, SG-Blend aims to harness their complementary strengths: SSwish's controlled non-monotonicity and symmetry, and GELU's smooth, probabilistic profile, to achieve a more universally robust balance between model expressivity and gradient stability. We conduct comprehensive empirical evaluations across diverse modalities and architectures, showing performance improvements across all considered natural language and computer vision tasks and models. These results, achieved with negligible computational overhead, underscore SG-Blend's potential as a versatile, drop-in replacement that consistently outperforms strong contemporary baselines. The code is available at https://anonymous.4open.science/r/SGBlend-6CBC.

We introduce EASG-Bench, a question-answering benchmark for egocentric videos where question-answer pairs are created from spatio-temporally grounded dynamic scene graphs that capture actions and the relationships among the camera wearer and objects. We propose a systematic evaluation framework and evaluate several language-only and video large language models (video-LLMs) on this benchmark. We observe a performance gap between language-only and video-LLMs, especially on questions focusing on temporal ordering, thereby identifying a research gap in the area of long-context video understanding. To promote the reproducibility of our findings and facilitate further research, the benchmark and accompanying code are available at the following GitHub page: https://github.com/fpv-iplab/EASGbench.

Pretraining egocentric vision-language models has become essential to improving downstream egocentric video-text tasks. These egocentric foundation models commonly use the transformer architecture. The memory footprint of these models during pretraining can be substantial. Therefore, we pretrain SViTT-Ego, the first sparse egocentric video-text transformer model integrating edge and node sparsification. We pretrain on the EgoClip dataset and incorporate the egocentric-friendly objective EgoNCE, instead of the frequently used InfoNCE. Most notably, SViTT-Ego obtains a +2.8% gain on EgoMCQ (intra-video) accuracy compared to LAVILA large, with no additional data augmentation techniques other than standard image augmentations, yet pretrainable on memory-limited devices.

We propose a graph-based representation learning framework for video summarization. First, we convert an input video to a graph where nodes correspond to each of the video frames. Then, we impose sparsity on the graph by connecting only those pairs of nodes that are within a specified temporal distance. We then formulate the video summarization task as a binary node classification problem, precisely classifying video frames whether they should belong to the output summary video. A graph constructed this way aims to capture long-range interactions among video frames, and the sparsity ensures the model trains without hitting the memory and compute bottleneck. Experiments on two datasets(SumMe and TVSum) demonstrate the effectiveness of the proposed nimble model compared to existing state-of-the-art summarization approaches while being one order of magnitude more efficient in compute time and memory

We present Egocentric Action Scene Graphs (EASGs), a new representation for long-form understanding of egocentric videos. EASGs extend standard manually-annotated representations of egocentric videos, such as verb-noun action labels, by providing a temporally evolving graph-based description of the actions performed by the camera wearer, including interacted objects, their relationships, and how actions unfold in time. Through a novel annotation procedure, we extend the Ego4D dataset by adding manually labeled Egocentric Action Scene Graphs offering a rich set of annotations designed for long-from egocentric video understanding. We hence define the EASG generation task and provide a baseline approach, establishing preliminary benchmarks. Experiments on two downstream tasks, egocentric action anticipation and egocentric activity summarization, highlight the effectiveness of EASGs for long-form egocentric video understanding. We will release the dataset and the code to replicate experiments and annotations.

We introduce LAVITI, a novel approach to learning language, video, and temporal representations in long-form videos via contrastive learning. Different from pre-training on video-text pairs like EgoVLP, LAVITI aims to align language, video, and temporal features by extracting meaningful moments in untrimmed videos. Our model employs a set of learnable moment queries to decode clip-level visual, language, and temporal features. In addition to vision and language alignment, we introduce relative temporal embeddings (TE) to represent timestamps in videos, which enables contrastive learning of time. Significantly different from traditional approaches, the prediction of a particular timestamp is transformed by computing the similarity score between the predicted TE and all TEs. Furthermore, existing approaches for video understanding are mainly designed for short videos due to high computational complexity and memory footprint. Our method can be trained on the Ego4D dataset with only 8 NVIDIA RTX-3090 GPUs in a day. We validated our method on CharadesEgo action recognition, achieving state-of-the-art results.

Do video-text transformers learn to model temporal relationships across frames? Despite their immense capacity and the abundance of multimodal training data, recent work has revealed the strong tendency of video-text models towards frame-based spatial representations, while temporal reasoning remains largely unsolved. In this work, we identify several key challenges in temporal learning of video-text transformers: the spatiotemporal trade-off from limited network size; the curse of dimensionality for multi-frame modeling; and the diminishing returns of semantic information by extending clip length. Guided by these findings, we propose SViTT, a sparse video-text architecture that performs multi-frame reasoning with significantly lower cost than naive transformers with dense attention. Analogous to graph-based networks, SViTT employs two forms of sparsity: edge sparsity that limits the query-key communications between tokens in self-attention, and node sparsity that discards uninformative visual tokens. Trained with a curriculum which increases model sparsity with the clip length, SViTT outperforms dense transformer baselines on multiple video-text retrieval and question answering benchmarks, with a fraction of computational cost. Project page: http://svcl.ucsd.edu/projects/svitt.

The task of dynamic scene graph generation (SGG) from videos is complicated and challenging due to the inherent dynamics of a scene, temporal fluctuation of model predictions, and the long-tailed distribution of the visual relationships in addition to the already existing challenges in image-based SGG. Existing methods for dynamic SGG have primarily focused on capturing spatio-temporal context using complex architectures without addressing the challenges mentioned above, especially the long-tailed distribution of relationships. This often leads to the generation of biased scene graphs. To address these challenges, we introduce a new framework called TEMPURA: TEmporal consistency and Memory Prototype guided UnceRtainty Attenuation for unbiased dynamic SGG. TEMPURA employs object-level temporal consistencies via transformer-based sequence modeling, learns to synthesize unbiased relationship representations using memory-guided training, and attenuates the predictive uncertainty of visual relations using a Gaussian Mixture Model (GMM). Extensive experiments demonstrate that our method achieves significant (up to 10% in some cases) performance gain over existing methods highlight- ing its superiority in generating more unbiased scene graphs. Code: https://github.com/sayaknag/unbiasedSGG.git