Search

Capitalizing on vast amount of image-text data, large-scale vision-language pre-training has demonstrated remarkable zero-shot capabilities and has been utilized in several applications. However, models trained on general everyday web-crawled data often exhibit sub-optimal performance for specialized domains, likely due to domain shift. Recent works have tackled this problem for some domains (e.g., healthcare) by constructing domain-specialized image-text data. However, constructing a dedicated large-scale image-text dataset for sustainable area of agriculture and livestock is still open to research. Further, this domain desires fine-grained feature learning due to the subtle nature of the downstream tasks (e.g, nutrient deficiency detection, livestock breed classification). To address this we present AgriCLIP, a vision-language foundational model dedicated to the domain of agriculture and livestock. First, we propose a large-scale dataset, named ALive, that leverages customized prompt generation strategy to overcome the scarcity of expert annotations. Our ALive dataset covers crops, livestock, and fishery, with around 600,000 image-text pairs. Second, we propose a training pipeline that integrates both contrastive and self-supervised learning to learn both global semantic and local fine-grained domain-specialized features. Experiments on diverse set of 20 downstream tasks demonstrate the effectiveness of AgriCLIP framework, achieving an absolute gain of 7.8\% in terms of average zero-shot classification accuracy, over the standard CLIP adaptation via domain-specialized ALive dataset. Our ALive dataset and code can be accessible at \href{https://github.com/umair1221/AgriCLIP/tree/main}{Github}.

Most recommender systems adopt collaborative filtering (CF) and provide recommendations based on past collective interactions. Therefore, the performance of CF algorithms degrades when few or no interactions are available, a scenario referred to as cold-start. To address this issue, previous work relies on models leveraging both collaborative data and side information on the users or items. Similar to multimodal learning, these models aim at combining collaborative and content representations in a shared embedding space. In this work we propose a novel technique for multimodal recommendation, relying on a multimodal Single-Branch embedding network for Recommendation (SiBraR). Leveraging weight-sharing, SiBraR encodes interaction data as well as multimodal side information using the same single-branch embedding network on different modalities. This makes SiBraR effective in scenarios of missing modality, including cold start. Our extensive experiments on large-scale recommendation datasets from three different recommendation domains (music, movie, and e-commerce) and providing multimodal content information (audio, text, image, labels, and interactions) show that SiBraR significantly outperforms CF as well as state-of-the-art content-based RSs in cold-start scenarios, and is competitive in warm scenarios. We show that SiBraR's recommendations are accurate in missing modality scenarios, and that the model is able to map different modalities to the same region of the shared embedding space, hence reducing the modality gap., Comment: Accepted at 18th ACM Conference on Recommender Systems (RecSys '24)

Strontium titanate is a classic quantum paraelectric oxide material that has been widely studied in bulk and thin films. It exhibits a well-known cubic-to-tetragonal antiferrodistortive phase transition at 105 K, characterized by the rotation of oxygen octahedra. A possible second phase transition at lower temperature is suppressed by quantum fluctuations, preventing the onset of ferroelectric order. However, recent studies have shown that ferroelectric order can be established at low temperatures by inducing strain and other means. Here, we used in situ multi-reflection Bragg coherent X-ray diffraction imaging to measure the strain and rotation tensors for two strontium titanate microcrystals at low temperature. We observe strains induced by dislocations and inclusion-like impurities in the microcrystals. Based on radial magnitude plots, these strains increase in magnitude and spread as the temperature decreases. Pearson's correlation heatmaps show a structural transition at 50 K, which we associate with the formation of a low-temperature ferroelectric phase in the presence of strain. We do not observe any change in local strains associated with the tetragonal phase transition at 105 K., Comment: 15 pages, 19 figures. Zenodo link will be active once published in a peer-reviewed journal

In this paper we compute genus 0 orbifold Gromov--Witten invariants of Calabi--Yau threefold complete intersections in weighted projective stacks, regardless of convexity conditions. The traditional quantumn Lefschetz principle may fail even for invariants with ambient insertions. Using quasimap wall-crossing, we are able to compute invariants with insertions from a specific subring of the Chen--Ruan cohomology, which contains all the ambient cohomology classes. Quasimap wall-crossing gives a mirror theorem expressing the I-function in terms of the J-function via a mirror map. The key of this paper is to find a suitable GIT presentation of the target space, so that the mirror map is invertible. An explicit formula for the I-function is given for all those target spaces and many examples with explicit computations of invariants are provided., Comment: 38 pgs; comments are welcome!

Multimodal networks have demonstrated remarkable performance improvements over their unimodal counterparts. Existing multimodal networks are designed in a multi-branch fashion that, due to the reliance on fusion strategies, exhibit deteriorated performance if one or more modalities are missing. In this work, we propose a modality invariant multimodal learning method, which is less susceptible to the impact of missing modalities. It consists of a single-branch network sharing weights across multiple modalities to learn inter-modality representations to maximize performance as well as robustness to missing modalities. Extensive experiments are performed on four challenging datasets including textual-visual (UPMC Food-101, Hateful Memes, Ferramenta) and audio-visual modalities (VoxCeleb1). Our proposed method achieves superior performance when all modalities are present as well as in the case of missing modalities during training or testing compared to the existing state-of-the-art methods.

Multimodal learning has demonstrated remarkable performance improvements over unimodal architectures. However, multimodal learning methods often exhibit deteriorated performances if one or more modalities are missing. This may be attributed to the commonly used multi-branch design containing modality-specific streams making the models reliant on the availability of a complete set of modalities. In this work, we propose a robust textual-visual multimodal learning method, Chameleon, that completely deviates from the conventional multi-branch design. To enable this, we present the unification of input modalities into one format by encoding textual modality into visual representations. As a result, our approach does not require modality-specific branches to learn modality-independent multimodal representations making it robust to missing modalities. Extensive experiments are performed on four popular challenging datasets including Hateful Memes, UPMC Food-101, MM-IMDb, and Ferramenta. Chameleon not only achieves superior performance when all modalities are present at train/test time but also demonstrates notable resilience in the case of missing modalities.

This paper is the first to present a novel, non-contact method that utilizes orthogonal frequency division multiplexing (OFDM) signals (of frequency 5.23 GHz, emitted by a software defined radio) to radio-expose the pulmonary patients in order to differentiate between five prevalent respiratory diseases, i.e., Asthma, Chronic obstructive pulmonary disease (COPD), Interstitial lung disease (ILD), Pneumonia (PN), and Tuberculosis (TB). The fact that each pulmonary disease leads to a distinct breathing pattern, and thus modulates the OFDM signal in a different way, motivates us to acquire OFDM-Breathe dataset, first of its kind. It consists of 13,920 seconds of raw RF data (at 64 distinct OFDM frequencies) that we have acquired from a total of 116 subjects in a hospital setting (25 healthy control subjects, and 91 pulmonary patients). Among the 91 patients, 25 have Asthma, 25 have COPD, 25 have TB, 5 have ILD, and 11 have PN. We implement a number of machine and deep learning models in order to do lung disease classification using OFDM-Breathe dataset. The vanilla convolutional neural network outperforms all the models with an accuracy of 97%, and stands out in terms of precision, recall, and F1-score. The ablation study reveals that it is sufficient to radio-observe the human chest on seven different microwave frequencies only, in order to make a reliable diagnosis (with 96% accuracy) of the underlying lung disease. This corresponds to a sensing overhead that is merely 10.93% of the allocated bandwidth. This points to the feasibility of 6G integrated sensing and communication (ISAC) systems of future where 89.07% of bandwidth still remains available for information exchange amidst on-demand health sensing. Through 6G ISAC, this work provides a tool for mass screening for respiratory diseases (e.g., COVID-19) at public places., Comment: submitted to a journal, 12 pages, 5 figures, 5 tables

Robots interacting with humans must be safe, reactive and adapt online to unforeseen environmental and task changes. Achieving these requirements concurrently is a challenge as interactive planners lack formal safety guarantees, while safe motion planners lack flexibility to adapt. To tackle this, we propose a modular control architecture that generates both safe and reactive motion plans for human-robot interaction by integrating temporal logic-based discrete task level plans with continuous Dynamical System (DS)-based motion plans. We formulate a reactive temporal logic formula that enables users to define task specifications through structured language, and propose a planning algorithm at the task level that generates a sequence of desired robot behaviors while being adaptive to environmental changes. At the motion level, we incorporate control Lyapunov functions and control barrier functions to compute stable and safe continuous motion plans for two types of robot behaviors: (i) complex, possibly periodic motions given by autonomous DS and (ii) time-critical tasks specified by Signal Temporal Logic~(STL). Our methodology is demonstrated on the Franka robot arm performing wiping tasks on a whiteboard and a mannequin that is compliant to human interactions and adaptive to environmental changes.

The advancements of technology have led to the use of multimodal systems in various real-world applications. Among them, the audio-visual systems are one of the widely used multimodal systems. In the recent years, associating face and voice of a person has gained attention due to presence of unique correlation between them. The Face-voice Association in Multilingual Environments (FAME) Challenge 2024 focuses on exploring face-voice association under a unique condition of multilingual scenario. This condition is inspired from the fact that half of the world's population is bilingual and most often people communicate under multilingual scenario. The challenge uses a dataset namely, Multilingual Audio-Visual (MAV-Celeb) for exploring face-voice association in multilingual environments. This report provides the details of the challenge, dataset, baselines and task details for the FAME Challenge., Comment: ACM Multimedia Conference - Grand Challenge

Currently, people with disability or difficulty to move their arms (referred to as "patients") have very limited technological solutions to efficiently address their physiological limitations. It is mainly due to two reasons: (1) the non-invasive solutions like mind-controlled prosthetic devices are typically very costly and require expensive maintenance; and (2) other solutions require costly invasive brain surgery, which is high risk to perform, expensive, and difficult to maintain. Therefore, current technological solutions are not accessible for all patients with different financial backgrounds. Toward this, we propose a low-cost technological solution called MindArm, a mechanized intelligent non-invasive neuro-driven prosthetic arm system. Our MindArm system employs a deep neural network (DNN) engine to translate brain signals into the intended prosthetic arm motion, thereby helping patients to perform many activities despite their physiological limitations. Here, our MindArm system utilizes widely accessible and low-cost surface electroencephalogram (EEG) electrodes coupled with an Open Brain Computer Interface and UDP networking for acquiring brain signals and transmitting them to the compute module for signal processing. In the compute module, we run a trained DNN model to interpret normalized micro-voltage of the brain signals, and then translate them into a prosthetic arm action via serial communication seamlessly. The experimental results on a fully working prototype demonstrate that, from the three defined actions, our MindArm system achieves positive success rates, i.e., 91\% for idle/stationary, 85\% for shake hand, and 84\% for pick-up cup. This demonstrates that our MindArm provides a novel approach for an alternate low-cost mind-controlled prosthetic devices for all patients., Comment: 8 pages, 21 figures, paper submitted to IROS 24, authors affiliated to NYUAD