Vision-Language Models (VLMs), with their powerful content generation capabilities, have been successfully applied to data annotation processes. However, the VLM-generated labels exhibit dual limitations: low quality (i.e., label noise) and absence of error correction mechanisms. To enhance label quality, we propose Human-Corrected Labels (HCLs), a novel setting that efficient human correction for VLM-generated noisy labels. As shown in Figure 1(b), HCL strategically deploys human correction only for instances with VLM discrepancies, achieving both higher-quality annotations and reduced labor costs. Specifically, we theoretically derive a risk-consistent estimator that incorporates both human-corrected labels and VLM predictions to train classifiers. Besides, we further propose a conditional probability method to estimate the label distribution using a combination of VLM outputs and model predictions. Extensive experiments demonstrate that our approach achieves superior classification performance and is robust to label noise, validating the effectiveness of HCL in practical weak supervision scenarios. Code https://github.com/Lilianach24/HCL.git

We propose a novel framework and a solution to tackle the continual learning (CL) problem with changing network architectures. Most CL methods focus on adapting a single architecture to a new task/class by modifying its weights. However, with rapid progress in architecture design, the problem of adapting existing solutions to novel architectures becomes relevant. To address this limitation, we propose Heterogeneous Continual Learning (HCL), where a wide range of evolving network architectures emerge continually together with novel data/tasks. As a solution, we build on top of the distillation family of techniques and modify it to a new setting where a weaker model takes the role of a teacher; meanwhile, a new stronger architecture acts as a student. Furthermore, we consider a setup of limited access to previous data and propose Quick Deep Inversion (QDI) to recover prior task visual features to support knowledge transfer. QDI significantly reduces computational costs compared to previous solutions and improves overall performance. In summary, we propose a new setup for CL with a modified knowledge distillation paradigm and design a quick data inversion method to enhance distillation. Our evaluation of various benchmarks shows a significant improvement on accuracy in comparison to state-of-the-art methods over various networks architectures.

As a combination of visual and audio signals, video is inherently multi-modal. However, existing video generation methods are primarily intended for the synthesis of visual frames, whereas audio signals in realistic videos are disregarded. In this work, we concentrate on a rarely investigated problem of text guided sounding video generation and propose the Sounding Video Generator (SVG), a unified framework for generating realistic videos along with audio signals. Specifically, we present the SVG-VQGAN to transform visual frames and audio melspectrograms into discrete tokens. SVG-VQGAN applies a novel hybrid contrastive learning method to model inter-modal and intra-modal consistency and improve the quantized representations. A cross-modal attention module is employed to extract associated features of visual frames and audio signals for contrastive learning. Then, a Transformer-based decoder is used to model associations between texts, visual frames, and audio signals at token level for auto-regressive sounding video generation. AudioSetCap, a human annotated text-video-audio paired dataset, is produced for training SVG. Experimental results demonstrate the superiority of our method when compared with existing textto-video generation methods as well as audio generation methods on Kinetics and VAS datasets.

Contrastive learning has emerged as a powerful tool for graph representation learning. However, most contrastive learning methods learn features of graphs with fixed coarse-grained scale, which might underestimate either local or global information. To capture more hierarchical and richer representation, we propose a novel Hierarchical Contrastive Learning (HCL) framework that explicitly learns graph representation in a hierarchical manner. Specifically, HCL includes two key components: a novel adaptive Learning to Pool (L2Pool) method to construct more reasonable multi-scale graph topology for more comprehensive contrastive objective, a novel multi-channel pseudo-siamese network to further enable more expressive learning of mutual information within each scale. Comprehensive experimental results show HCL achieves competitive performance on 12 datasets involving node classification, node clustering and graph classification. In addition, the visualization of learned representation reveals that HCL successfully captures meaningful characteristics of graphs.

While showing impressive performance on various kinds of learning tasks, it is yet unclear whether deep learning models have the ability to robustly tackle reasoning tasks. than by learning the underlying reasoning process that is actually required to solve the tasks. Measuring the robustness of reasoning in machine learning models is challenging as one needs to provide a task that cannot be easily shortcut by exploiting spurious statistical correlations in the data, while operating on complex objects and constraints. reasoning task. To address this issue, we propose ChemAlgebra, a benchmark for measuring the reasoning capabilities of deep learning models through the prediction of stoichiometrically-balanced chemical reactions. ChemAlgebra requires manipulating sets of complex discrete objects -- molecules represented as formulas or graphs -- under algebraic constraints such as the mass preservation principle. We believe that ChemAlgebra can serve as a useful test bed for the next generation of machine reasoning models and as a promoter of their development.

Digitalisation is a strategic priority for public healthcare policy: major innovations in medicine are expected from the development of IT tools and the generalisation of digital models, which will contribute to improving the efficiency of healthcare systems and services, writes Inria in this press release.. In order to guide the digital transition of this sector in the Lyon region, Inria and the Hospices Civils de Lyon (HCL), the second largest hospital centre in France after AP-HP (Paris), are pooling their expertise to create a centre for the development of artificial intelligence and a joint project team dedicated to digital models for neuroscience. On July 2021, the two bodies signed a "memorandum of understanding", prior to concluding a framework agreement for tripartite collaboration (Inria / HCL / Claude Bernard–Lyon 1 University), thus officialising the details of this shared ambition. "The signing of a large-scale partnership between Lyon Public Hospitals and Inria testifies to the firm positioning of the University Hospital and its academic partners with regard to digital healthcare", says Raymond Le Moign, Managing Director of Hospices Civils de Lyon. "For the last ten years, the application of digital science in the healthcare sector has been a key focal point for Inria. Almost a third of its teams lead research which opens opportunities in this field", according to Hugues Berry, senior researcher in molecular neuroscience and Deputy Scientific Director at Inria.

Emotion recognition in conversation (ERC) aims to detect the emotion label for each utterance. Motivated by recent studies which have proven that feeding training examples in a meaningful order rather than considering them randomly can boost the performance of models, we propose an ERC-oriented hybrid curriculum learning framework. Our framework consists of two curricula: (1) conversation-level curriculum (CC); and (2) utterance-level curriculum (UC). In CC, we construct a difficulty measurer based on "emotion shift" frequency within a conversation, then the conversations are scheduled in an "easy to hard" schema according to the difficulty score returned by the difficulty measurer. For UC, it is implemented from an emotion-similarity perspective, which progressively strengthens the model's ability in identifying the confusing emotions. With the proposed model-agnostic hybrid curriculum learning strategy, we observe significant performance boosts over a wide range of existing ERC models and we are able to achieve new state-of-the-art results on four public ERC datasets.

HCL Technologies (HCL), a leading global technology company, today announced it has joined NVIDIA Partner Network (NPN), the US-based technology company's partner program, with plans to pursue expanded opportunities in the Artificial Intelligence (AI) space. HCL also announced the launch of NEXT.ai, HCL's AI Lab, focused on upskilling engineers and architects utilizing AI solutions across hardware and software technology stacks. This will enhance their capabilities to build AI-led solutions for multiple HCL clients across industry verticals. The HCL NEXT.ai lab focuses on emerging technologies to develop industry-specific AI solutions for customers pursuing digital transformation.

I was in Boston last week for the HCL Technologies advisor and analyst event, which occurred on the day the world woke up to a new US political direction. President elect Trump's stated ambitions to restrict outsourcing gave many food for thought at the event, but HCL's modern approaches to digital business arguably render old outsourcing models to the history books anyway. I thought HCL presented a very well organized understanding of the enterprise scale challenges large business entities face, solutioning approaches and ways they partner with their clients. The healthy appetite they have for sharing risks and rewards in a rapidly accelerating and evolving business world spoke volumes about understanding changing business strategies and associated digital technology deployments. 'Vijay' detailed significant new shifts in the ways HCL does business.

New Delhi: Former HCL Technologies Ltd chief executive officer (CEO) Anant Gupta on Friday launched a Rs100 crore investment fund TECHCELX, to invest in start-ups focused on digital technologies such as machine learning, Internet of Things, artificial intelligence and analytics. HCL on Friday announced the elevation of its chief operating officer C. Vijayakumar to the top position in the company to replace Gupta. TECHCELX will invest between Rs50 lakh and Rs10 crore in disruptive technology ventures and the shareholding will range between 15% and 51%. "With TECHCELX, I am pursuing my long-term goals and passion towards promoting our country's entrepreneurship and innovation potential in the technology space," said Gupta, chairman and CEO, TECHCELX. Three other key partners from investment banking and technology sectors are backing this fund, the company said in a statement.