Hateful memes detection is a challenging multimodal understanding task that requires comprehensive learning of vision, language, and cross-modal interactions. Previous research has focused on developing effective fusion strategies for integrating hate information from different modalities. However, these methods excessively rely on cross-modal fusion features, ignoring the modality uncertainty caused by the contribution degree of each modality to hate sentiment and the modality imbalance caused by the dominant modality suppressing the optimization of another modality. To this end, this paper proposes an Uncertainty-guided Modal Rebalance (UMR) framework for hateful memes detection. The uncertainty of each meme is explicitly formulated by designing stochastic representation drawn from a Gaussian distribution for aggregating cross-modal features with unimodal features adaptively. The modality imbalance is alleviated by improving cosine loss from the perspectives of inter-modal feature and weight vectors constraints. In this way, the suppressed unimodal representation ability in multimodal models would be unleashed, while the learning of modality contribution would be further promoted. Extensive experimental results demonstrate that the proposed UMR produces the state-of-the-art performance on four widely-used datasets.

Uncertainty-Guided Modal Rebalance for Hateful Memes Detection

Transformer-based methods have gone mainstream in multimodal sequential learning. The intra and inter modality interactions are captured by the query-key associations of multi-head attention. In this way, the calculated multimodal contexts (attentional results) are expected to be relevant to the query modality. However, in existing literature, the alignment degree between different calculated attentional results of the same query are under-explored. Based on this concern, we propose a new constrained scheme called Multimodal Contextual Contrast (MCC), which could align the multiple attentional results from both local and global perspectives, making the information capture more efficient. Concretely, the calculated attentional results of different modalities are mapped into a common feature space, those attentional vectors with the same query are considered as a positive group and the remaining sets are negative. From local perspective, we sample the negative groups for a positive group by randomly changing the sequential step of one specific context and keeping the other stay the same. From coarse global perspective, we divide all the contextual groups into two sets (i.e., aligned and unaligned), making the total score of aligned group relatively large. We extend the vectorial inner product operation for more input and calculate the aligned score for each multimodal group. Considering that the computational complexity scales exponentially to the number of modalities, we adopt stochastic expectation approximation (SEA) for the real process. The extensive experimental results on several tasks reveal the effectiveness of our contributions.

Rethinking the Multimodal Correlation of Multimodal Sequential Learning via Generalizable Attentional Results Alignment

Vision-and-language (VL) models with separate encoders for each modality (e.g., CLIP) have become the go-to models for zero-shot image classification and image-text retrieval. They are, however, mostly evaluated in English as multilingual benchmarks are limited in availability. We introduce Babel-ImageNet, a massively multilingual benchmark that offers (partial) translations of ImageNet labels to 100 languages, built without machine translation or manual annotation. We instead automatically obtain reliable translations by linking them -- via shared WordNet synsets -- to BabelNet, a massively multilingual lexico-semantic network. We evaluate 11 public multilingual CLIP models on zero-shot image classification (ZS-IC) on our benchmark, demonstrating a significant gap between English ImageNet performance and that of high-resource languages (e.g., German or Chinese), and an even bigger gap for low-resource languages (e.g., Sinhala or Lao). Crucially, we show that the models' ZS-IC performance highly correlates with their performance in image-text retrieval, validating the use of Babel-imageNet to evaluate multilingual models for the vast majority of languages without gold image-text data. Finally, we show that the performance of multilingual CLIP can be drastically improved for low-resource languages with parameter-efficient language-specific training. 
We make our code and data publicly available: \url{https://github.com/gregor-ge/Babel-ImageNet}

Babel-ImageNet: Massively Multilingual Evaluation of Vision-and-Language Representations

Multi-change captioning aims to describe complex and coupled changes within an image pair in natural language. Compared with single-change captioning, this task requires the model to have higher-level cognition ability to reason an arbitrary number of changes. In this paper, we propose a novel context-aware difference distilling (CARD) network to capture all genuine changes for yielding sentences. Given an image pair, CARD first decouples context features that aggregate all similar/dissimilar semantics, termed common/difference context features. Then, the consistency and independence constraints are designed to guarantee the alignment/discrepancy of common/difference context features. Further, the common context features guide the model to mine locally unchanged features, which are subtracted from the pair to distill locally difference features. Next, the difference context features augment the locally difference features to ensure that all changes are distilled. In this way, we obtain an omni-representation of all changes, which is translated into linguistic sentences by a transformer decoder. Extensive experiments on three public datasets show CARD performs favourably against state-of-the-art methods. The code is available at https://github.com/tuyunbin/CARD.

Context-aware Difference Distilling for Multi-change Captioning

In the field of speech synthesis, there is a growing emphasis on employing multimodal speech to enhance robustness. A key challenge in this area is the scarcity of datasets that pair audio with corresponding video. We employ a methodology that incorporates modality alignment during the pre-training phase on multimodal datasets, uniquely facilitating zero-shot generalization through the process of freezing the video modality feature extraction component and the encoder module within the pretrained weights, thereby enabling effective cross-modal and cross-lingual transfer. We have named this method `Uni-Dubbing'. Our method finely tunes with both multimodal and single-modality audio data. In multimodal scenarios, it achieves a reduced word error rate (WER) of 31.73\%, surpassing the previous best of 33.9\%. It also excels in metrics like tone quality and synchronization. With single-modality audio, it achieves a WER of 36.08\%, demonstrating adaptability to limited data. Its domain generalization capabilities are proven across various language tasks in video translation and audio generation. Trained on 433 hours of audio data, it surpasses techniques using 200 hours of audiovisual data. The code and demo are available at \href{https://diracer.github.io/unidubbing}{https://diracer.github.io/unidubbing}.

Uni-Dubbing: Zero-Shot Speech Synthesis from Visual Articulation

Recent generalist vision-language models (VLMs) have demonstrated impressive reasoning capabilities across diverse multimodal tasks. However, these models still struggle with fine-grained object-level understanding and grounding. In terms of modeling, existing VLMs implicitly align text tokens with image patch tokens, which is ineffective for embedding alignment at the same granularity and inevitably introduces noisy spurious background features. Additionally, these models struggle when generalizing to unseen visual concepts and may not be reliable for domain-specific tasks without further fine-tuning. To address these limitations, we propose a novel method to prompt large language models with in-context visual object vectors, thereby enabling controllable object-level reasoning. This eliminates the necessity of fusing a lengthy array of image patch features and significantly speeds up training. Furthermore, we propose region-level retrieval using our object representations, facilitating rapid adaptation to new objects without additional training. Our experiments reveal that our method achieves competitive referring object classification and captioning performance, while also offering zero-shot generalization and robustness to visually challenging contexts.

OLIVE: Object Level In-Context Visual Embeddings

The goal of this work is to create a robotic subsystem that is capable of effectively fighting fires in a shipboard environment both autonomously and via teleoperation. The system will be designed as a firefighting attachment for mobile robots incorporated into the fleet in the future. SAFERS will use a COTS motorized firefighting monitor, a variable power aimed water cannon, to target and extinguish fires. The targeting control software will allow for both automatic targeting of the fire and manual targeting control by an operator. Both automatic and manual targeting decisions will be informed by the various cameras that make up the perception system. The perception system will also use the video feeds to identify fires and hotspots within the fires to provide ideal targeting suggestions to most effectively fight fires. A variety of different perception sensors will be explored in order to test their effectiveness for firefighting tasks in a complex environment. Additionally, various learning and classical vision algorithms will be trained and tested to evaluate performance in the tasks of fire and hot spot detection across sensor spectrums. Some of the machine learning methods currently being explored are deep convolutional neural networks and visual transformers for object detection, semantic segmentation, and monocular depth estimation. The detected objects and estimated depths of the environments will be used by the control algorithms to make intelligent targeting decisions during firefighting scenarios. Ideally, the developed system will be useful tool for fleet firefighters as both an autonomous and remote system that will lower risk to sailors while simultaneously improving overall firefighting capability.



Semi-Autonomous Fire Extinguishing Robot Subsystem (SAFERS)

Naval Asset Corrosion and Defect Segmentation Using UNet Architecture 

Abstract: 

Corrosion poses a significant challenge in the maintenance and upkeep of naval assets, leading to degradation and high cost. Timely and accurate detection of corrosion and other defects are crucial to ensure operational efficiency and system availability. Traditional methods for detection often involve manual inspections, which are time-consuming, prone to human error, and require inspectors with specialized training and expertise. 

There is a significant amount of research into automatic detection of corrosion via computer vision algorithms and semantic segmentation that we utilized to combat this problem. Our method makes use of hand-annotated high-resolution images of existing weathered structures (conex boxes) as inputs for a machine learning model to detect basic corrosion, as well as classify additional defects such as paint blistering and cracking. Due to the limited amount of data representing the extended classes, image augmentations were performed to expand our dataset. 

When classifying “stuff”, where the defects can be any shape or size, having an extremely high level of precision is necessary. To account for this, a UNet machine learning architecture, which consists of a contracting and symmetric expanding path, was chosen due to its capability of performing precise pixel-level image segmentation with fewer training data requirements. We have experimented with multiple variations of UNet and implemented additional modules like attention and deep supervision in order to increase the overall performance and generalization of the model.

Naval Asset Corrosion and Defect Segmentation Using UNet Architecture 

Our task is to create an equivalent to “follow me” mode for ships at sea. Allowing a Sailor to launch a small quadcopter and forget about it while it follows and circles the ship of interest could be incredibly useful. Our particular application is autonomous inspection and battle damage assessment, say in the aftermath of a Naval casualty. 

We develop a computer vision pipeline that uses a drone’s camera feed to track the behavior of a ship. It uses camera frames to estimate the ship’s position and orientation to potentially correct the drone’s trajectory, ensuring that the ship is always in view, even in adverse conditions. Our proposed method can also be effective in broader applications in fields like Autonomous Driving and Robotics, where the main agent takes an action based on its surrounding environment. 

The first stage of our computer vision pipeline consists of calculating the orientation of the ship. Our method passes the images from the camera through a UNet image segmentation model to get the ship’s mask and boundaries. We then use an OpenCV algorithm to generate a rotated rectangle that minimizes its area while enclosing the entire mask of the ship. This rotated bounding box effectively represents the direction that the ship is facing based on its physical features, so we use its angle to estimate the ship’s orientation with respect to the drone. This technique is beneficial because it is less complex than 3D Pose Estimation models, allowing for faster real-time processing of image data. 

For our image segmentation model to perform effectively in various environmental conditions, it is crucial to have a variety of flight paths, elevation angles, and visibility conditions to prevent overfitting. Collecting real data with these criteria is an expensive and laborious process, so we have started with Unity-generated synthetic data with relevant variations as ground truth labels for training. Using this data to train our computer vision model resulted in highly accurate predicted masks from unseen flight test images.

Estimating Ship Position and Orientation using Deep Neural Networks and Synthetic Data

The objective of this line of research is to use Convolutional Neural
Networks (CNNs) to automatically identify corrosion in photographs.
This approach provides a reliable alternative to traditional
approaches. Corrosion, a widespread issue in several sectors,
necessitates prompt detection to avoid structural harm and guarantee
operational security. The use of deep learning, particularly
Convolutional Neural Networks (CNNs), offers an advanced method for
detecting corrosion. This is achieved by using the neural networks'
capability to identify complex patterns in picture data. This research
investigates the creation and assessment of a Convolutional Neural
Network (CNN) model that was trained using a dataset consisting of
photos depicting both corrosion and non-corrosion. The model's
effectiveness is assessed by analyzing key performance indicators such
as True Positive Rate, False Positive Rate, accuracy, recall, and area
under the ROC curve. The study illuminates the capabilities and
constraints of the CNN-based technique, providing useful insights for
businesses in need of effective and precise corrosion detection
solutions.


Corrosion Detection Using Deep Learning

Detection of fast micro UAVs using event-based cameras must be insensitive to polarity effects. This was revealed by using polarity offset and contrast maximization as methods to detect motion from event-based data.

Computing Motion from Events-Based Cameras for UAV Surveillance

Deep learning models such as CNNs have surpassed human
performance in computer vision tasks such as image classi-
fication. However, despite their sophistication, these models
lack interpretability which can lead to biased outcomes re-
flecting existing prejudices in the data. We aim to make pre-
dictions made by a CNN interpretable. Hence, we present a
novel framework called NeSyFOLD to create a neurosym-
bolic (NeSy) model for image classification tasks. The model
is a CNN with all layers following the last convolutional layer
replaced by a stratified answer set program (ASP) derived
from the last layer kernels. The answer set program can be
viewed as a rule-set, wherein the truth value of each pred-
icate depends on the activation of the corresponding kernel
in the CNN. The rule-set serves as a global explanation for
the model and is interpretable. We also use our NeSyFOLD
framework with a CNN that is trained using a sparse kernel
learning technique called Elite BackProp (EBP). This leads to
a significant reduction in rule-set size without compromising
accuracy or fidelity thus improving scalability of the NeSy
model and interpretability of its rule-set. Evaluation is done
on datasets with varied complexity and sizes. We also pro-
pose a novel algorithm for labelling the predicates in the rule-
set with meaningful semantic concept(s) learnt by the CNN.
We evaluate the performance of our “semantic labelling algo-
rithm” to quantify the efficacy of the semantic labelling for
both the NeSy model and the NeSy-EBP model.

Underline digital video library

Uncertainty-Guided Modal Rebalance for Hateful Memes Detection

Rethinking the Multimodal Correlation of Multimodal Sequential Learning via Generalizable Attentional Results Alignment

Babel-ImageNet: Massively Multilingual Evaluation of Vision-and-Language Representations

Context-aware Difference Distilling for Multi-change Captioning

Uni-Dubbing: Zero-Shot Speech Synthesis from Visual Articulation

OLIVE: Object Level In-Context Visual Embeddings

Semi-Autonomous Fire Extinguishing Robot Subsystem (SAFERS)

Naval Asset Corrosion and Defect Segmentation Using UNet Architecture

Estimating Ship Position and Orientation using Deep Neural Networks and Synthetic Data

Corrosion Detection Using Deep Learning

Computing Motion from Events-Based Cameras for UAV Surveillance

NeSyFOLD: A Framework for Interpretable Image Classification

Stay up to date with the latest Underline news!

PRODUCT

COMPANY

RESOURCES