Vivek Kumar

Lijun Yu, Yong Cheng, Zhiruo Wang et al. · cmu, deepmind

In this work, we introduce Semantic Pyramid AutoEncoder (SPAE) for enabling frozen LLMs to perform both understanding and generation tasks involving non-linguistic modalities such as images or videos. SPAE converts between raw pixels and interpretable lexical tokens (or words) extracted from the LLM's vocabulary. The resulting tokens capture both the semantic meaning and the fine-grained details needed for visual reconstruction, effectively translating the visual content into a language comprehensible to the LLM, and empowering it to perform a wide array of multimodal tasks. Our approach is validated through in-context learning experiments with frozen PaLM 2 and GPT 3.5 on a diverse set of image understanding and generation tasks. Our method marks the first successful attempt to enable a frozen LLM to generate image content while surpassing state-of-the-art performance in image understanding tasks, under the same setting, by over 25%.

Jennifer J. Sun, Markus Marks, Andrew Ulmer et al.

We introduce MABe22, a large-scale, multi-agent video and trajectory benchmark to assess the quality of learned behavior representations. This dataset is collected from a variety of biology experiments, and includes triplets of interacting mice (4.7 million frames video+pose tracking data, 10 million frames pose only), symbiotic beetle-ant interactions (10 million frames video data), and groups of interacting flies (4.4 million frames of pose tracking data). Accompanying these data, we introduce a panel of real-life downstream analysis tasks to assess the quality of learned representations by evaluating how well they preserve information about the experimental conditions (e.g. strain, time of day, optogenetic stimulation) and animal behavior. We test multiple state-of-the-art self-supervised video and trajectory representation learning methods to demonstrate the use of our benchmark, revealing that methods developed using human action datasets do not fully translate to animal datasets. We hope that our benchmark and dataset encourage a broader exploration of behavior representation learning methods across species and settings.

Vivek Kumar, Rishabh Maheshwary, Vikram Pudi

Existing Math Word Problem (MWP) solvers have achieved high accuracy on benchmark datasets. However, prior works have shown that such solvers do not generalize well and rely on superficial cues to achieve high performance. In this paper, we first conduct experiments to showcase that this behaviour is mainly associated with the limited size and diversity present in existing MWP datasets. Next, we propose several data augmentation techniques broadly categorized into Substitution and Paraphrasing based methods. By deploying these methods we increase the size of existing datasets by five folds. Extensive experiments on two benchmark datasets across three state-of-the-art MWP solvers show that proposed methods increase the generalization and robustness of existing solvers. On average, proposed methods significantly increase the state-of-the-art results by over five percentage points on benchmark datasets. Further, the solvers trained on the augmented dataset perform comparatively better on the challenge test set. We also show the effectiveness of proposed techniques through ablation studies and verify the quality of augmented samples through human evaluation.

Vivek Kumar, Sushmita Singh, Prayag Tiwari

Our system, VISU, participated in the WASSA 2023 Shared Task (3) of Emotion Classification from essays written in reaction to news articles. Emotion detection from complex dialogues is challenging and often requires context/domain understanding. Therefore in this research, we have focused on developing deep learning (DL) models using the combination of word embedding representations with tailored prepossessing strategies to capture the nuances of emotions expressed. Our experiments used static and contextual embeddings (individual and stacked) with Bidirectional Long short-term memory (BiLSTM) and Transformer based models. We occupied rank tenth in the emotion detection task by scoring a Macro F1-Score of 0.2717, validating the efficacy of our implemented approaches for small and imbalanced datasets with mixed categories of target emotions.

Tom Gunter, Zirui Wang, Chong Wang et al.

We present foundation language models developed to power Apple Intelligence features, including a ~3 billion parameter model designed to run efficiently on devices and a large server-based language model designed for Private Cloud Compute. These models are designed to perform a wide range of tasks efficiently, accurately, and responsibly. This report describes the model architecture, the data used to train the model, the training process, how the models are optimized for inference, and the evaluation results. We highlight our focus on Responsible AI and how the principles are applied throughout the model development.

Narayana Darapaneni, Ashish Ranjan, Dany Bright et al.

With the advancement in AI, deep learning techniques are widely used to design robust classification models in several areas such as medical diagnosis tasks in which it achieves good performance. In this paper, we have proposed the CNN model (Convolutional Neural Network) for the classification of Chest X-ray images for Radiological Society of North America Pneumonia (RSNA) datasets. The study also tries to achieve the same RSNA benchmark results using the limited computational resources by trying out various approaches to the methodologies that have been implemented in recent years. The proposed method is based on a non-complex CNN and the use of transfer learning algorithms like Xception, InceptionV3/V4, EfficientNetB7. Along with this, the study also tries to achieve the same RSNA benchmark results using the limited computational resources by trying out various approaches to the methodologies that have been implemented in recent years. The RSNA benchmark MAP score is 0.25, but using the Mask RCNN model on a stratified sample of 3017 along with image augmentation gave a MAP score of 0.15. Meanwhile, the YoloV3 without any hyperparameter tuning gave the MAP score of 0.32 but still, the loss keeps decreasing. Running the model for a greater number of iterations can give better results.

Ethan Li, Anders Boesen Lindbo Larsen, Chen Zhang et al. · apple-ml, cmu

We introduce two multilingual, multimodal foundation language models that power Apple Intelligence features across Apple devices and services: i a 3B-parameter on-device model optimized for Apple silicon through architectural innovations such as KV-cache sharing and 2-bit quantization-aware training; and ii a scalable server model built on a novel Parallel-Track Mixture-of-Experts PT-MoE transformer that combines track parallelism, mixture-of-experts sparse computation, and interleaved global-local attention to deliver high quality with competitive cost on Apple's Private Cloud Compute platform. Both models are trained on large-scale multilingual and multimodal datasets sourced via responsible web crawling, licensed corpora, and high-quality synthetic data, then further refined with supervised fine-tuning and reinforcement learning on a new asynchronous platform. The resulting models support several additional languages while understanding images and executing tool calls. In public benchmarks and human evaluations, both the server model and the on-device model match or surpass comparably sized open baselines. A new Swift-centric Foundation Models framework exposes guided generation, constrained tool calling, and LoRA adapter fine-tuning, allowing developers to integrate these capabilities with a few lines of code. The latest advancements in Apple Intelligence models are grounded in our Responsible AI approach with safeguards like content filtering and locale-specific evaluation, as well as our commitment to protecting our users' privacy with innovations like Private Cloud Compute.

Aneesh Panchal, Kirti Beniwal, Vivek Kumar

Nonlinear mathematical models introduce the relation between various physical and biological interactions present in nature. One of the most famous models is the Lotka-Volterra model which defined the interaction between predator and prey species present in nature. However, predators, scavengers, and prey populations coexist in a natural system where scavengers can additionally rely on the dead bodies of predators present in the system. Keeping this in mind, the formulation and simulation of the predator prey scavenger model is introduced in this paper. For the predation response, respective prey species are assumed to have Holling's functional response of type III. The proposed model is tested for various simulations and is found to be showing satisfactory results in different scenarios. After simulations, the American forest dataset is taken for parameter estimation which imitates the real-world case. For parameter estimation, a physics-informed deep neural network is used with the Adam backpropagation method which prevents the avalanche effect in trainable parameters updation. For neural networks, mean square error and physics-informed informed error are considered. After the neural network, the hence-found parameters are fine-tuned using the Broyden-Fletcher-Goldfarb-Shanno algorithm. Finally, the hence-found parameters using a natural dataset are tested for stability using Jacobian stability analysis. Future research work includes minimization of error induced by parameters, bifurcation analysis, and sensitivity analysis of the parameters.

Vivek Kumar, Eirini Ntoutsi, Pushpraj Singh Rajawat et al.

Large language models (LLMs) have shown promising capabilities in healthcare analysis but face several challenges like hallucinations, parroting, and bias manifestation. These challenges are exacerbated in complex, sensitive, and low-resource domains. Therefore, in this work we introduce IC-AnnoMI, an expert-annotated motivational interviewing (MI) dataset built upon AnnoMI by generating in-context conversational dialogues leveraging LLMs, particularly ChatGPT. IC-AnnoMI employs targeted prompts accurately engineered through cues and tailored information, taking into account therapy style (empathy, reflection), contextual relevance, and false semantic change. Subsequently, the dialogues are annotated by experts, strictly adhering to the Motivational Interviewing Skills Code (MISC), focusing on both the psychological and linguistic dimensions of MI dialogues. We comprehensively evaluate the IC-AnnoMI dataset and ChatGPT's emotional reasoning ability and understanding of domain intricacies by modeling novel classification tasks employing several classical machine learning and current state-of-the-art transformer approaches. Finally, we discuss the effects of progressive prompting strategies and the impact of augmented data in mitigating the biases manifested in IC-AnnoM. Our contributions provide the MI community with not only a comprehensive dataset but also valuable insights for using LLMs in empathetic text generation for conversational therapy in supervised settings.

Artem Dementyev, Wazeer Zulfikar, Sinan Hersek et al.

Current multimodal LLMs process audio as a mono stream, ignoring the rich spatial information essential for embodied AI. Existing spatial audio models, conversely, are constrained to fixed microphone geometries, preventing deployment across diverse devices. We present PhaseCoder, a transformer-only spatial audio encoder that is agnostic to microphone geometry. PhaseCoder takes raw multichannel audio and microphone coordinates as inputs to perform localization and produces robust spatial embeddings. We demonstrate that Gemma 3n LLM can be fine-tuned to reason over "Spatial Audio Tokens" produced by PhaseCoder. We show our encoder achieves state-of-the-art results on microphone-invariant localization benchmarks and, for the first time, enables an LLM to perform complex spatial reasoning and targeted transcription tasks from an arbitrary microphone array.

Vivek Kumar, Pushpraj Singh Rajawat, Eirini Ntoutsi

Recent advancements in large language models (LLMs) have shown their potential across both general and domain-specific tasks. However, there is a growing concern regarding their lack of sensitivity, factual incorrectness in responses, inconsistent expressions of empathy, bias, hallucinations, and overall inability to capture the depth and complexity of human understanding, especially in low-resource and sensitive domains such as psychology. To address these challenges, our study employs a mixed-methods approach to evaluate the efficacy of LLMs in psychotherapy. We use LLMs to generate precise summaries of motivational interviewing (MI) dialogues and design a two-stage annotation scheme based on key components of the Motivational Interviewing Treatment Integrity (MITI) framework, namely evocation, collaboration, autonomy, direction, empathy, and a non-judgmental attitude. Using expert-annotated MI dialogues as ground truth, we formulate multi-class classification tasks to assess model performance under progressive prompting techniques, incorporating one-shot and few-shot prompting. Our results offer insights into LLMs' capacity for understanding complex psychological constructs and highlight best practices to mitigate ``semantic drift" in therapeutic settings. Our work contributes not only to the MI community by providing a high-quality annotated dataset to address data scarcity in low-resource domains but also critical insights for using LLMs for precise contextual interpretation in complex behavioral therapy.

Vivek Kumar, Himanshu Sahu, Hari Prabhat Gupta et al.

Yoga is a discipline of physical postures, breathing techniques, and meditative practices rooted in ancient Indian traditions, now embraced worldwide for promoting overall well-being and inner balance. The practices are a large set of items, our term for executable actions like physical poses or breath exercises, to offer for a person's well-being. However, to get benefits of Yoga tailored to a person's unique needs, a person needs to (a) discover their subset from the large and seemingly complex set with inter-dependencies, (b) continue to follow them with interest adjusted to their changing abilities and near-term objectives, and (c) as appropriate, adapt to alternative items based on changing environment and the person's health conditions. In this vision paper, we describe the challenges for the Yoga personalization problem. Next, we sketch a preliminary approach and use the experience to provide an outlook on solving the challenging problem using existing and novel techniques from a multidisciplinary computing perspective. To the best of our knowledge, this is the first paper that comprehensively examines decision support issues around Yoga personalization, from pose sensing to recommendation of corrections for a complete regimen, and illustrates with a case study of Surya Namaskar -- a set of 12 choreographed poses.

Vadim Liventsev, Vivek Kumar, Allmin Pradhap Singh Susaiyah et al.

The use of machine learning in Healthcare has the potential to improve patient outcomes as well as broaden the reach and affordability of Healthcare. The history of other application areas indicates that strong benchmarks are essential for the development of intelligent systems. We present Personal Health Interfaces Leveraging HUman-MAchine Natural interactions (PhilHumans), a holistic suite of benchmarks for machine learning across different Healthcare settings - talk therapy, diet coaching, emergency care, intensive care, obstetric sonography - as well as different learning settings, such as action anticipation, timeseries modeling, insight mining, language modeling, computer vision, reinforcement learning and program synthesis

Simone Balloccu, Ehud Reiter, Vivek Kumar et al.

Large Language Models (LLMs), with their flexible generation abilities, can be powerful data sources in domains with few or no available corpora. However, problems like hallucinations and biases limit such applications. In this case study, we pick nutrition counselling, a domain lacking any public resource, and show that high-quality datasets can be gathered by combining LLMs, crowd-workers and nutrition experts. We first crowd-source and cluster a novel dataset of diet-related issues, then work with experts to prompt ChatGPT into producing related supportive text. Finally, we let the experts evaluate the safety of the generated text. We release HAI-coaching, the first expert-annotated nutrition counselling dataset containing ~2.4K dietary struggles from crowd workers, and ~97K related supportive texts generated by ChatGPT. Extensive analysis shows that ChatGPT while producing highly fluent and human-like text, also manifests harmful behaviours, especially in sensitive topics like mental health, making it unsuitable for unsupervised use.

Vivek Kumar, Rishabh Maheshwary, Vikram Pudi

Standard accuracy metrics have shown that Math Word Problem (MWP) solvers have achieved high performance on benchmark datasets. However, the extent to which existing MWP solvers truly understand language and its relation with numbers is still unclear. In this paper, we generate adversarial attacks to evaluate the robustness of state-of-the-art MWP solvers. We propose two methods Question Reordering and Sentence Paraphrasing to generate adversarial attacks. We conduct experiments across three neural MWP solvers over two benchmark datasets. On average, our attack method is able to reduce the accuracy of MWP solvers by over 40 percentage points on these datasets. Our results demonstrate that existing MWP solvers are sensitive to linguistic variations in the problem text. We verify the validity and quality of generated adversarial examples through human evaluation.

Zixiu Wu, Rim Helaoui, Vivek Kumar et al.

Empathetic response from the therapist is key to the success of clinical psychotherapy, especially motivational interviewing. Previous work on computational modelling of empathy in motivational interviewing has focused on offline, session-level assessment of therapist empathy, where empathy captures all efforts that the therapist makes to understand the client's perspective and convey that understanding to the client. In this position paper, we propose a novel task of turn-level detection of client need for empathy. Concretely, we propose to leverage pre-trained language models and empathy-related general conversation corpora in a unique labeller-detector framework, where the labeller automatically annotates a motivational interviewing conversation corpus with empathy labels to train the detector that determines the need for therapist empathy. We also lay out our strategies of extending the detector with additional-input and multi-task setups to improve its detection and explainability.

Danilo Dessi, Rim Helaoui, Vivek Kumar et al.

Today, we are seeing an ever-increasing number of clinical notes that contain clinical results, images, and textual descriptions of patient's health state. All these data can be analyzed and employed to cater novel services that can help people and domain experts with their common healthcare tasks. However, many technologies such as Deep Learning and tools like Word Embeddings have started to be investigated only recently, and many challenges remain open when it comes to healthcare domain applications. To address these challenges, we propose the use of Deep Learning and Word Embeddings for identifying sixteen morbidity types within textual descriptions of clinical records. For this purpose, we have used a Deep Learning model based on Bidirectional Long-Short Term Memory (LSTM) layers which can exploit state-of-the-art vector representations of data such as Word Embeddings. We have employed pre-trained Word Embeddings namely GloVe and Word2Vec, and our own Word Embeddings trained on the target domain. Furthermore, we have compared the performances of the deep learning approaches against the traditional tf-idf using Support Vector Machine and Multilayer perceptron (our baselines). From the obtained results it seems that the latter outperforms the combination of Deep Learning approaches using any word embeddings. Our preliminary results indicate that there are specific features that make the dataset biased in favour of traditional machine learning approaches.

Berkan Kadioglu, Michael Horgan, Xiaoyu Liu et al.

Conv-TasNet is a recently proposed waveform-based deep neural network that achieves state-of-the-art performance in speech source separation. Its architecture consists of a learnable encoder/decoder and a separator that operates on top of this learned space. Various improvements have been proposed to Conv-TasNet. However, they mostly focus on the separator, leaving its encoder/decoder as a (shallow) linear operator. In this paper, we conduct an empirical study of Conv-TasNet and propose an enhancement to the encoder/decoder that is based on a (deep) non-linear variant of it. In addition, we experiment with the larger and more diverse LibriTTS dataset and investigate the generalization capabilities of the studied models when trained on a much larger dataset. We propose cross-dataset evaluation that includes assessing separations from the WSJ0-2mix, LibriTTS and VCTK databases. Our results show that enhancements to the encoder/decoder can improve average SI-SNR performance by more than 1 dB. Furthermore, we offer insights into the generalization capabilities of Conv-TasNet and the potential value of improvements to the encoder/decoder.

Manuel Mazzara, Ilya Afanasyev, Smruti R. Sarangi et al.

The vision encompassing Smart and Software-defined Buildings (SSDB) is becoming more and more popular and its implementation is now more accessible due to the widespread adoption of the IoT infrastructure. Some of the most important applications sustaining this vision are energy management, environmental comfort, safety and surveillance. This paper surveys IoT and SSB technologies and their cooperation towards the realization of Smart Spaces. We propose a four-layer reference architecture and we organize related concepts around it. This conceptual frame is useful to identify the current literature on the topic and to connect the dots into a coherent vision of the future of residential and commercial buildings.

Vivek Kumar, Brojo Kishore Mishra, Manuel Mazzara et al.

As much as data science is playing a pivotal role everywhere, healthcare also finds it prominent application. Breast Cancer is the top rated type of cancer amongst women; which took away 627,000 lives alone. This high mortality rate due to breast cancer does need attention, for early detection so that prevention can be done in time. As a potential contributor to state-of-art technology development, data mining finds a multi-fold application in predicting Brest cancer. This work focuses on different classification techniques implementation for data mining in predicting malignant and benign breast cancer. Breast Cancer Wisconsin data set from the UCI repository has been used as experimental dataset while attribute clump thickness being used as an evaluation class. The performances of these twelve algorithms: Ada Boost M 1, Decision Table, J Rip, Lazy IBK, Logistics Regression, Multiclass Classifier, Multilayer Perceptron, Naive Bayes, Random forest and Random Tree are analyzed on this data set. Keywords- Data Mining, Classification Techniques, UCI repository, Breast Cancer, Classification Algorithms

Vivek Kumar, Atul Samadhiya

Noise is an important factor which when get added to an image reduces its quality and appearance. So in order to enhance the image qualities, it has to be removed with preserving the textural information and structural features of image. There are different types of noises exist who corrupt the images. Selection of the denoising algorithm is application dependent. Hence, it is necessary to have knowledge about the noise present in the image so as to select the appropriate denoising algorithm. Objective of this paper is to present brief account on types of noises, its types and different noise removal algorithms. In the first section types of noises on the basis of their additive and multiplicative nature are being discussed. In second section a precise classification and analysis of the different potential image denoising algorithm is presented. At the end of paper, a comparative study of all these algorithms in context of performance evaluation is done and concluded with several promising directions for future research work.

Vivek Kumar, Atul Samadhiya

When an image is formed, factors such as lighting (spectra, source, and intensity) and camera characteristics (sensor response, lenses) affect the appearance of the image. Therefore, the prime factor that reduces the quality of the image is noise. It hides the important details and information of images. In order to enhance the qualities of the image, the removal of noises become imperative and that should not at the cost of any loss of image information. Noise removal is one of the pre-processing stages of image processing. In this paper a new method for the enhancement of grayscale images is introduced, when images are corrupted by fixed valued impulse noise (salt and pepper noise). The proposed methodology ensures a better output for the low and medium density of fixed value impulse noise as compared to the other famous filters like Standard Median Filter (SMF), Decision Based Median Filter (DBMF) and Modified Decision Based Median Filter (MDBMF) etc. The main objective of the proposed method was to improve peak signal to noise ratio (PSNR), visual perception and reduction in the blurring of the image. The proposed algorithm replaced the noisy pixel by trimmed mean value. When previous pixel values, 0s, and 255s are present in the particular window and all the pixel values are 0s and 255s then the remaining noisy pixels are replaced by mean value. The gray-scale image of mandrill and Lena were tested via the proposed method. The experimental result shows better peak signal to noise ratio (PSNR), mean square error values with better visual and human perception.

Vivek Kumar, Manuel Mazzara, Maj. Gen. et al.

Terrorism has become one of the most tedious problems to deal with and a prominent threat to mankind. To enhance counter-terrorism, several research works are developing efficient and precise systems, data mining is not an exception. Immense data is floating in our lives, though the scarce availability of authentic terrorist attack data in the public domain makes it complicated to fight terrorism. This manuscript focuses on data mining classification techniques and discusses the role of United Nations in counter-terrorism. It analyzes the performance of classifiers such as Lazy Tree, Multilayer Perceptron, Multiclass and Naïve Bayes classifiers for observing the trends for terrorist attacks around the world. The database for experiment purpose is created from different public and open access sources for years 1970-2015 comprising of 156,772 reported attacks causing massive losses of lives and property. This work enumerates the losses occurred, trends in attack frequency and places more prone to it, by considering the attack responsibilities taken as evaluation class.

Cong Zhou, Michael Horgan, Vivek Kumar et al.

Here we present a novel approach to conditioning the SampleRNN generative model for voice conversion (VC). Conventional methods for VC modify the perceived speaker identity by converting between source and target acoustic features. Our approach focuses on preserving voice content and depends on the generative network to learn voice style. We first train a multi-speaker SampleRNN model conditioned on linguistic features, pitch contour, and speaker identity using a multi-speaker speech corpus. Voice-converted speech is generated using linguistic features and pitch contour extracted from the source speaker, and the target speaker identity. We demonstrate that our system is capable of many-to-many voice conversion without requiring parallel data, enabling broad applications. Subjective evaluation demonstrates that our approach outperforms conventional VC methods.

Vivek Kumar, Sandeep Sharma

Cryptompress, a new 128-bit (initial) private-key cryptography algorithm is proposed. It uses a block size of at least 30 bits and increments prior key size to additional 32 bits on each unsuccessful attempt of any means, including bruteforcing, further changing a specific portion of the cyphertext using the reformed Feistel network. Encryption process results from a proposed compression sequence developed using lookup table and shift operations followed by key generation. Eventually, four matrixes named add-sub matrix, reduced matrix, sequence matrix and term matrix are obtained which ultimately forms a cyphertext.

Vivek Kumar, Sumit Pandey, Amrindra Pal et al.

Image recognition is the need of the hour. In order to be able to recognize an image, it is of immense importance that the image should be distinguishable from the background. In the present work, an approach is presented for automatic detection and recognition of regular 2D shapes in low noise environments. The work has a large number of direct applications in the real world. The algorithm proposed is based on locating the edges and thus, in turn calculating the area of the object helps in identification of a specified shape. The results were simulated using MATLAB tool are encouraging and validate the proposed algorithm. Index Terms: Edge Detection, Area Calculation, Shape Detection, Object Recognition