Preference modeling techniques, such as direct preference optimization (DPO), has shown effective in enhancing the generalization abilities of large language model (LLM). However, in tasks involving video instruction-following, providing informative feedback, especially for detecting hallucinations in generated responses, remains a significant challenge. Previous studies have explored using large large multimodal models (LMMs) as reward models to guide preference modeling, but their ability to accurately assess the factuality of generated responses compared to corresponding videos has not been conclusively established. This paper introduces a novel framework that utilizes detailed video captions as a proxy of video content, enabling language models to incorporate this information as supporting evidence for scoring video Question Answering (QA) predictions. Our approach demonstrates robust alignment with OpenAI GPT-4V model's reward mechanism, which directly takes video frames as input. Furthermore, we show that applying this tailored reward through DPO significantly improves the performance of video LMMs on video QA tasks.
Recent research in representation learning has shown that hierarchical data lends itself to low-dimensional and highly informative representations in hyperbolic space. However, even if hyperbolic embeddings have gathered attention in image recognition, their optimization is prone to numerical hurdles. Further, it remains unclear which applications stand to benefit the most from the implicit bias imposed by hyperbolicity, when compared to traditional Euclidean features. In this paper, we focus on prototypical hyperbolic neural networks. In particular, the tendency of hyperbolic embeddings to converge to the boundary of the Poincar\'e ball in high dimensions and the effect this has on few-shot classification. We show that the best few-shot results are attained for hyperbolic embeddings at a common hyperbolic radius. In contrast to prior benchmark results, we demonstrate that better performance can be achieved by a fixed-radius encoder equipped with the Euclidean metric, regardless of the embedding dimension.
We study the problem of human action recognition using motion capture (MoCap) sequences. Unlike existing techniques that take multiple manual steps to derive standardized skeleton representations as model input, we propose a novel Spatial-Temporal Mesh Transformer (STMT) to directly model the mesh sequences. The model uses a hierarchical transformer with intra-frame off-set attention and inter-frame self-attention. The attention mechanism allows the model to freely attend between any two vertex patches to learn non-local relationships in the spatial-temporal domain. Masked vertex modeling and future frame prediction are used as two self-supervised tasks to fully activate the bi-directional and auto-regressive attention in our hierarchical transformer. The proposed method achieves state-of-the-art performance compared to skeleton-based and point-cloud-based models on common MoCap benchmarks. Code is available at https://github.com/zgzxy001/STMT.
Grounded Situation Recognition (GSR) aims to generate structured semantic summaries of images for "human-like" event understanding. Specifically, GSR task not only detects the salient activity verb (e.g. buying), but also predicts all corresponding semantic roles (e.g. agent and goods). Inspired by object detection and image captioning tasks, existing methods typically employ a two-stage framework: 1) detect the activity verb, and then 2) predict semantic roles based on the detected verb. Obviously, this illogical framework constitutes a huge obstacle to semantic understanding. First, pre-detecting verbs solely without semantic roles inevitably fails to distinguish many similar daily activities (e.g., offering and giving, buying and selling). Second, predicting semantic roles in a closed auto-regressive manner can hardly exploit the semantic relations among the verb and roles. To this end, in this paper we propose a novel two-stage framework that focuses on utilizing such bidirectional relations within verbs and roles. In the first stage, instead of pre-detecting the verb, we postpone the detection step and assume a pseudo label, where an intermediate representation for each corresponding semantic role is learned from images. In the second stage, we exploit transformer layers to unearth the potential semantic relations within both verbs and semantic roles. With the help of a set of support images, an alternate learning scheme is designed to simultaneously optimize the results: update the verb using nouns corresponding to the image, and update nouns using verbs from support images. Extensive experimental results on challenging SWiG benchmarks show that our renovated framework outperforms other state-of-the-art methods under various metrics.
This paper studies zero-shot cross-lingual transfer of vision-language models. Specifically, we focus on multilingual text-to-video search and propose a Transformer-based model that learns contextualized multilingual multimodal embeddings. Under a zero-shot setting, we empirically demonstrate that performance degrades significantly when we query the multilingual text-video model with non-English sentences. To address this problem, we introduce a multilingual multimodal pre-training strategy, and collect a new multilingual instructional video dataset (MultiHowTo100M) for pre-training. Experiments on VTT show that our method significantly improves video search in non-English languages without additional annotations. Furthermore, when multilingual annotations are available, our method outperforms recent baselines by a large margin in multilingual text-to-video search on VTT and VATEX; as well as in multilingual text-to-image search on Multi30K. Our model and Multi-HowTo100M is available at http://github.com/berniebear/Multi-HT100M.
This paper studies how to introduce viewpoint-invariant feature representations that can help action recognition and detection. Although we have witnessed great progress of action recognition in the past decade, it remains challenging yet interesting how to efficiently model the geometric variations in large scale datasets. This paper proposes a novel Spatial-Temporal Alignment Network (STAN) that aims to learn geometric invariant representations for action recognition and action detection. The STAN model is very light-weighted and generic, which could be plugged into existing action recognition models like ResNet3D and the SlowFast with a very low extra computational cost. We test our STAN model extensively on AVA, Kinetics-400, AVA-Kinetics, Charades, and Charades-Ego datasets. The experimental results show that the STAN model can consistently improve the state of the arts in both action detection and action recognition tasks. We will release our data, models and code.
Social media has become an important tool to share information about crisis events such as natural disasters and mass attacks. Detecting actionable posts that contain useful information requires rapid analysis of huge volume of data in real-time. This poses a complex problem due to the large amount of posts that do not contain any actionable information. Furthermore, the classification of information in real-time systems requires training on out-of-domain data, as we do not have any data from a new emerging crisis. Prior work focuses on models pre-trained on similar event types. However, those models capture unnecessary event-specific biases, like the location of the event, which affect the generalizability and performance of the classifiers on new unseen data from an emerging new event. In our work, we train an adversarial neural model to remove latent event-specific biases and improve the performance on tweet importance classification.
The dominant paradigm for learning video-text representations -- noise contrastive learning -- increases the similarity of the representations of pairs of samples that are known to be related, such as text and video from the same sample, and pushes away the representations of all other pairs. We posit that this last behaviour is too strict, enforcing dissimilar representations even for samples that are semantically-related -- for example, visually similar videos or ones that share the same depicted action. In this paper, we propose a novel method that alleviates this by leveraging a generative model to naturally push these related samples together: each sample's caption must be reconstructed as a weighted combination of other support samples' visual representations. This simple idea ensures that representations are not overly-specialized to individual samples, are reusable across the dataset, and results in representations that explicitly encode semantics shared between samples, unlike noise contrastive learning. Our proposed method outperforms others by a large margin on MSR-VTT, VATEX and ActivityNet, for video-to-text and text-to-video retrieval.