Faruk Ahmed

Alexander H. Liu, Alexis Tacnet, Andy Ehrenberg et al. · deepmind, tsinghua

We introduce Voxtral TTS, an expressive multilingual text-to-speech model that generates natural speech from as little as 3 seconds of reference audio. Voxtral TTS adopts a hybrid architecture that combines auto-regressive generation of semantic speech tokens with flow-matching for acoustic tokens. These tokens are encoded and decoded with Voxtral Codec, a speech tokenizer trained from scratch with a hybrid VQ-FSQ quantization scheme. In human evaluations conducted by native speakers, Voxtral TTS is preferred for multilingual voice cloning due to its naturalness and expressivity, achieving a 68.4\% win rate over ElevenLabs Flash v2.5. We release the model weights under a CC BY-NC license.

Alexander H. Liu, Kartik Khandelwal, Sandeep Subramanian et al.

We introduce the Ministral 3 series, a family of parameter-efficient dense language models designed for compute and memory constrained applications, available in three model sizes: 3B, 8B, and 14B parameters. For each model size, we release three variants: a pretrained base model for general-purpose use, an instruction finetuned, and a reasoning model for complex problem-solving. In addition, we present our recipe to derive the Ministral 3 models through Cascade Distillation, an iterative pruning and continued training with distillation technique. Each model comes with image understanding capabilities, all under the Apache 2.0 license.

Jeremy Lai, Faruk Ahmed, Supriya Vijay et al.

Task-specific deep learning models in histopathology offer promising opportunities for improving diagnosis, clinical research, and precision medicine. However, development of such models is often limited by availability of high-quality data. Foundation models in histopathology that learn general representations across a wide range of tissue types, diagnoses, and magnifications offer the potential to reduce the data, compute, and technical expertise necessary to develop task-specific deep learning models with the required level of model performance. In this work, we describe the development and evaluation of foundation models for histopathology via self-supervised learning (SSL). We first establish a diverse set of benchmark tasks involving 17 unique tissue types and 12 unique cancer types and spanning different optimal magnifications and task types. Next, we use this benchmark to explore and evaluate histopathology-specific SSL methods followed by further evaluation on held out patch-level and weakly supervised tasks. We found that standard SSL methods thoughtfully applied to histopathology images are performant across our benchmark tasks and that domain-specific methodological improvements can further increase performance. Our findings reinforce the value of using domain-specific SSL methods in pathology, and establish a set of high quality foundation models to enable further research across diverse applications.

Faruk Ahmed, Md. Taimur Ahad, Yousuf Rayhan Emon

Tea leaf diseases are a major challenge to agricultural productivity, with far-reaching implications for yield and quality in the tea industry. The rise of machine learning has enabled the development of innovative approaches to combat these diseases. Early detection and diagnosis are crucial for effective crop management. For predicting tea leaf disease, several automated systems have already been developed using different image processing techniques. This paper delivers a systematic review of the literature on machine learning methodologies applied to diagnose tea leaf disease via image classification. It thoroughly evaluates the strengths and constraints of various Vision Transformer models, including Inception Convolutional Vision Transformer (ICVT), GreenViT, PlantXViT, PlantViT, MSCVT, Transfer Learning Model & Vision Transformer (TLMViT), IterationViT, IEM-ViT. Moreover, this paper also reviews models like Dense Convolutional Network (DenseNet), Residual Neural Network (ResNet)-50V2, YOLOv5, YOLOv7, Convolutional Neural Network (CNN), Deep CNN, Non-dominated Sorting Genetic Algorithm (NSGA-II), MobileNetv2, and Lesion-Aware Visual Transformer. These machine-learning models have been tested on various datasets, demonstrating their real-world applicability. This review study not only highlights current progress in the field but also provides valuable insights for future research directions in the machine learning-based detection and classification of tea leaf diseases.

Md Taimur Ahad, Sumaya Mustofa, Faruk Ahmed et al.

In deep learning, transfer learning and ensemble models have shown promise in improving computer-aided disease diagnosis. However, applying the transfer learning and ensemble model is still relatively limited. Moreover, the ensemble model's development is ad-hoc, overlooks redundant layers, and suffers from imbalanced datasets and inadequate augmentation. Lastly, significant Deep Convolutional Neural Networks (D-CNNs) have been introduced to detect and classify breast cancer. Still, very few comparative studies were conducted to investigate the accuracy and efficiency of existing CNN architectures. Realising the gaps, this study compares the performance of D-CNN, which includes the original CNN, transfer learning, and an ensemble model, in detecting breast cancer. The comparison study of this paper consists of comparison using six CNN-based deep learning architectures (SE-ResNet152, MobileNetV2, VGG19, ResNet18, InceptionV3, and DenseNet-121), a transfer learning, and an ensemble model on breast cancer detection. Among the comparison of these models, the ensemble model provides the highest detection and classification accuracy of 99.94% for breast cancer detection and classification. However, this study also provides a negative result in the case of transfer learning, as the transfer learning did not increase the accuracy of the original SE-ResNet152, MobileNetV2, VGG19, ResNet18, InceptionV3, and DenseNet-121 model. The high accuracy in detecting and categorising breast cancer detection using CNN suggests that the CNN model is promising in breast cancer disease detection. This research is significant in biomedical engineering, computer-aided disease diagnosis, and ML-based disease detection.

Gheorghe Comanici, Eric Bieber, Mike Schaekermann et al. · amazon-science, baidu

In this report, we introduce the Gemini 2.X model family: Gemini 2.5 Pro and Gemini 2.5 Flash, as well as our earlier Gemini 2.0 Flash and Flash-Lite models. Gemini 2.5 Pro is our most capable model yet, achieving SoTA performance on frontier coding and reasoning benchmarks. In addition to its incredible coding and reasoning skills, Gemini 2.5 Pro is a thinking model that excels at multimodal understanding and it is now able to process up to 3 hours of video content. Its unique combination of long context, multimodal and reasoning capabilities can be combined to unlock new agentic workflows. Gemini 2.5 Flash provides excellent reasoning abilities at a fraction of the compute and latency requirements and Gemini 2.0 Flash and Flash-Lite provide high performance at low latency and cost. Taken together, the Gemini 2.X model generation spans the full Pareto frontier of model capability vs cost, allowing users to explore the boundaries of what is possible with complex agentic problem solving.

Theo Guidroz, Diego Ardila, Jimmy Li et al.

Information on the web, such as scientific publications and Wikipedia, often surpasses users' reading level. To help address this, we used a self-refinement approach to develop a LLM capability for minimally lossy text simplification. To validate our approach, we conducted a randomized study involving 4563 participants and 31 texts spanning 6 broad subject areas: PubMed (biomedical scientific articles), biology, law, finance, literature/philosophy, and aerospace/computer science. Participants were randomized to viewing original or simplified texts in a subject area, and answered multiple-choice questions (MCQs) that tested their comprehension of the text. The participants were also asked to provide qualitative feedback such as task difficulty. Our results indicate that participants who read the simplified text answered more MCQs correctly than their counterparts who read the original text (3.9% absolute increase, p<0.05). This gain was most striking with PubMed (14.6%), while more moderate gains were observed for finance (5.5%), aerospace/computer science (3.8%) domains, and legal (3.5%). Notably, the results were robust to whether participants could refer back to the text while answering MCQs. The absolute accuracy decreased by up to ~9% for both original and simplified setups where participants could not refer back to the text, but the ~4% overall improvement persisted. Finally, participants' self-reported perceived ease based on a simplified NASA Task Load Index was greater for those who read the simplified text (absolute change on a 5-point scale 0.33, p<0.05). This randomized study, involving an order of magnitude more participants than prior works, demonstrates the potential of LLMs to make complex information easier to understand. Our work aims to enable a broader audience to better learn and make use of expert knowledge available on the web, improving information accessibility.

Faruk Ahmed, Lin Yang, Tiam Jaroensri et al.

The interpretation of histopathology cases underlies many important diagnostic and treatment decisions in medicine. Notably, this process typically requires pathologists to integrate and summarize findings across multiple slides per case. Existing vision-language capabilities in computational pathology have so far been largely limited to small regions of interest, larger regions at low magnification, or single whole-slide images (WSIs). This limits interpretation of findings that span multiple high-magnification regions across multiple WSIs. By making use of Gemini 1.5 Flash, a large multimodal model (LMM) with a 1-million token context window, we demonstrate the ability to generate bottom-line diagnoses from up to 40,000 768x768 pixel image patches from multiple WSIs at 10X magnification. This is the equivalent of up to 11 hours of video at 1 fps. Expert pathologist evaluations demonstrate that the generated report text is clinically accurate and equivalent to or preferred over the original reporting for 68% (95% CI: [60%, 76%]) of multi-slide examples with up to 5 slides. While performance decreased for examples with 6 or more slides, this study demonstrates the promise of leveraging the long-context capabilities of modern LMMs for the uniquely challenging task of medical report generation where each case can contain thousands of image patches.

Atilla P. Kiraly, Sebastien Baur, Kenneth Philbrick et al.

Robust medical Machine Learning (ML) models have the potential to revolutionize healthcare by accelerating clinical research, improving workflows and outcomes, and producing novel insights or capabilities. Developing such ML models from scratch is cost prohibitive and requires substantial compute, data, and time (e.g., expert labeling). To address these challenges, we introduce Health AI Developer Foundations (HAI-DEF), a suite of pre-trained, domain-specific foundation models, tools, and recipes to accelerate building ML for health applications. The models cover various modalities and domains, including radiology (X-rays and computed tomography), histopathology, dermatological imaging, and audio. These models provide domain specific embeddings that facilitate AI development with less labeled data, shorter training times, and reduced computational costs compared to traditional approaches. In addition, we utilize a common interface and style across these models, and prioritize usability to enable developers to integrate HAI-DEF efficiently. We present model evaluations across various tasks and conclude with a discussion of their application and evaluation, covering the importance of ensuring efficacy, fairness, and equity. Finally, while HAI-DEF and specifically the foundation models lower the barrier to entry for ML in healthcare, we emphasize the importance of validation with problem- and population-specific data for each desired usage setting. This technical report will be updated over time as more modalities and features are added.

Yousuf Rayhan Emon, Md Golam Rabbani, Md. Taimur Ahad et al.

Sweet orange leaf diseases are significant to agricultural productivity. Leaf diseases impact fruit quality in the citrus industry. The apparition of machine learning makes the development of disease finder. Early detection and diagnosis are necessary for leaf management. Sweet orange leaf disease-predicting automated systems have already been developed using different image-processing techniques. This comprehensive literature review is systematically based on leaf disease and machine learning methodologies applied to the detection of damaged leaves via image classification. The benefits and limitations of different machine learning models, including Vision Transformer (ViT), Neural Network (CNN), CNN with SoftMax and RBF SVM, Hybrid CNN-SVM, HLB-ConvMLP, EfficientNet-b0, YOLOv5, YOLOv7, Convolutional, Deep CNN. These machine learning models tested on various datasets and detected the disease. This comprehensive review study related to leaf disease compares the performance of the models; those models' accuracy, precision, recall, etc., were used in the subsisting studies

Alexander H. Liu, Andy Ehrenberg, Andy Lo et al.

We introduce Voxtral Realtime, a natively streaming automatic speech recognition model that matches offline transcription quality at sub-second latency. Unlike approaches that adapt offline models through chunking or sliding windows, Voxtral Realtime is trained end-to-end for streaming, with explicit alignment between audio and text streams. Our architecture builds on the Delayed Streams Modeling framework, introducing a new causal audio encoder and Ada RMS-Norm for improved delay conditioning. We scale pretraining to a large-scale dataset spanning 13 languages. At a delay of 480ms, Voxtral Realtime achieves performance on par with Whisper, the most widely deployed offline transcription system. We release the model weights under the Apache 2.0 license.

Jalal Ahmmed, Faruk Ahmed, Rashedul Hasan Shohan et al.

Mango is an important fruit crop in South Asia, but its cultivation is frequently hampered by leaf diseases that greatly impact yield and quality. This research examines the performance of five pre-trained convolutional neural networks, DenseNet201, InceptionV3, ResNet152V2, SeResNet152, and Xception, for multi-class identification of mango leaf diseases across eight classes using a transfer learning strategy with fine-tuning. The models were assessed through standard evaluation metrics, such as accuracy, precision, recall, F1-score, and confusion matrices. Among the architectures tested, DenseNet201 delivered the best results, achieving 99.33% accuracy with consistently strong metrics for individual classes, particularly excelling in identifying Cutting Weevil and Bacterial Canker. Moreover, ResNet152V2 and SeResNet152 provided strong outcomes, whereas InceptionV3 and Xception exhibited lower performance in visually similar categories like Sooty Mould and Powdery Mildew. The training and validation plots demonstrated stable convergence for the highest-performing models. The capability of fine-tuned transfer learning models, for precise and dependable multi-class mango leaf disease detection in intelligent agricultural applications.

Saifuddin Sagor, Md Taimur Ahad, Faruk Ahmed et al.

Early and accurate detection through Pap smear analysis is critical to improving patient outcomes and reducing mortality of Cervical cancer. State-of-the-art (SOTA) Convolutional Neural Networks (CNNs) require substantial computational resources, extended training time, and large datasets. In this study, a lightweight CNN model, S-Net (Simple Net), is developed specifically for cervical cancer detection and classification using Pap smear images to address these limitations. Alongside S-Net, six SOTA CNNs were evaluated using transfer learning, including multi-path (DenseNet201, ResNet152), depth-based (Serasnet152), width-based multi-connection (Xception), depth-wise separable convolutions (MobileNetV2), and spatial exploitation-based (VGG19). All models, including S-Net, achieved comparable accuracy, with S-Net reaching 99.99%. However, S-Net significantly outperforms the SOTA CNNs in terms of computational efficiency and inference time, making it a more practical choice for real-time and resource-constrained applications. A major limitation in CNN-based medical diagnosis remains the lack of transparency in the decision-making process. To address this, Explainable AI (XAI) techniques, such as SHAP, LIME, and Grad-CAM, were employed to visualize and interpret the key image regions influencing model predictions. The novelty of this study lies in the development of a highly accurate yet computationally lightweight model (S-Net) caPable of rapid inference while maintaining interpretability through XAI integration. Furthermore, this work analyzes the behavior of SOTA CNNs, investigates the effects of negative transfer learning on Pap smear images, and examines pixel intensity patterns in correctly and incorrectly classified samples.

Vishwesh Ramanathan, Tony Xu, Pushpak Pati et al.

Prediction tasks in digital pathology are challenging due to the massive size of whole-slide images (WSIs) and the weak nature of training signals. Advances in computing, data availability, and self-supervised learning (SSL) have paved the way for slide-level foundation models (SLFMs) that can improve prediction tasks in low-data regimes. However, current methods under-utilize shared information between tasks and modalities. To overcome this challenge, we propose ModalTune, a novel fine-tuning framework which introduces the Modal Adapter to integrate new modalities without modifying SLFM weights. Additionally, we use large-language models (LLMs) to encode labels as text, capturing semantic relationships across multiple tasks and cancer types in a single training recipe. ModalTune achieves state-of-the-art (SOTA) results against both uni-modal and multi-modal models across four cancer types, jointly improving survival and cancer subtype prediction while remaining competitive in pan-cancer settings. Additionally, we show ModalTune is generalizable to two out-of-distribution (OOD) datasets. To our knowledge, this is the first unified fine-tuning framework for multi-modal, multi-task, and pan-cancer modeling in digital pathology.

Faruk Ahmed, Andrew Sellergren, Lin Yang et al.

Microscopic interpretation of histopathology images underlies many important diagnostic and treatment decisions. While advances in vision-language modeling raise new opportunities for analysis of such images, the gigapixel-scale size of whole slide images (WSIs) introduces unique challenges. Additionally, pathology reports simultaneously highlight key findings from small regions while also aggregating interpretation across multiple slides, often making it difficult to create robust image-text pairs. As such, pathology reports remain a largely untapped source of supervision in computational pathology, with most efforts relying on region-of-interest annotations or self-supervision at the patch-level. In this work, we develop a vision-language model based on the BLIP-2 framework using WSIs paired with curated text from pathology reports. This enables applications utilizing a shared image-text embedding space, such as text or image retrieval for finding cases of interest, as well as integration of the WSI encoder with a frozen large language model (LLM) for WSI-based generative text capabilities such as report generation or AI-in-the-loop interactions. We utilize a de-identified dataset of over 350,000 WSIs and diagnostic text pairs, spanning a wide range of diagnoses, procedure types, and tissue types. We present pathologist evaluation of text generation and text retrieval using WSI embeddings, as well as results for WSI classification and workflow prioritization (slide-level triaging). Model-generated text for WSIs was rated by pathologists as accurate, without clinically significant error or omission, for 78% of WSIs on average. This work demonstrates exciting potential capabilities for language-aligned WSI embeddings.

Lin Yang, Shawn Xu, Andrew Sellergren et al.

Many clinical tasks require an understanding of specialized data, such as medical images and genomics, which is not typically found in general-purpose large multimodal models. Building upon Gemini's multimodal models, we develop several models within the new Med-Gemini family that inherit core capabilities of Gemini and are optimized for medical use via fine-tuning with 2D and 3D radiology, histopathology, ophthalmology, dermatology and genomic data. Med-Gemini-2D sets a new standard for AI-based chest X-ray (CXR) report generation based on expert evaluation, exceeding previous best results across two separate datasets by an absolute margin of 1% and 12%, where 57% and 96% of AI reports on normal cases, and 43% and 65% on abnormal cases, are evaluated as "equivalent or better" than the original radiologists' reports. We demonstrate the first ever large multimodal model-based report generation for 3D computed tomography (CT) volumes using Med-Gemini-3D, with 53% of AI reports considered clinically acceptable, although additional research is needed to meet expert radiologist reporting quality. Beyond report generation, Med-Gemini-2D surpasses the previous best performance in CXR visual question answering (VQA) and performs well in CXR classification and radiology VQA, exceeding SoTA or baselines on 17 of 20 tasks. In histopathology, ophthalmology, and dermatology image classification, Med-Gemini-2D surpasses baselines across 18 out of 20 tasks and approaches task-specific model performance. Beyond imaging, Med-Gemini-Polygenic outperforms the standard linear polygenic risk score-based approach for disease risk prediction and generalizes to genetically correlated diseases for which it has never been trained. Although further development and evaluation are necessary in the safety-critical medical domain, our results highlight the potential of Med-Gemini across a wide range of medical tasks.

Samuel Lavoie, Faruk Ahmed, Aaron Courville

While unsupervised domain translation (UDT) has seen a lot of success recently, we argue that mediating its translation via categorical semantic features could broaden its applicability. In particular, we demonstrate that categorical semantics improves the translation between perceptually different domains sharing multiple object categories. We propose a method to learn, in an unsupervised manner, categorical semantic features (such as object labels) that are invariant of the source and target domains. We show that conditioning the style encoder of unsupervised domain translation methods on the learned categorical semantics leads to a translation preserving the digits on MNIST$\leftrightarrow$SVHN and to a more realistic stylization on Sketches$\to$Reals.

Faruk Ahmed, Md Sultan Mahmud, Kazi Ashraf Moinuddin et al.

Finding a path free from obstacles that poses minimal risk is critical for safe navigation. People who are sighted and people who are visually impaired require navigation safety while walking on a sidewalk. In this research we developed an assistive navigation on a sidewalk by integrating sensory inputs using reinforcement learning. We trained a Sidewalk Obstacle Avoidance Agent (SOAA) through reinforcement learning in a simulated robotic environment. A Sidewalk Obstacle Conversational Agent (SOCA) is built by training a natural language conversation agent with real conversation data. The SOAA along with SOCA was integrated in a prototype device called augmented guide (AG). Empirical analysis showed that this prototype improved the obstacle avoidance experience about 5% from a base case of 81.29%

Faruk Ahmed, Aaron Courville

We critically appraise the recent interest in out-of-distribution (OOD) detection and question the practical relevance of existing benchmarks. While the currently prevalent trend is to consider different datasets as OOD, we argue that out-distributions of practical interest are ones where the distinction is semantic in nature for a specified context, and that evaluative tasks should reflect this more closely. Assuming a context of object recognition, we recommend a set of benchmarks, motivated by practical applications. We make progress on these benchmarks by exploring a multi-task learning based approach, showing that auxiliary objectives for improved semantic awareness result in improved semantic anomaly detection, with accompanying generalization benefits.

Ishaan Gulrajani, Faruk Ahmed, Martin Arjovsky et al.

Generative Adversarial Networks (GANs) are powerful generative models, but suffer from training instability. The recently proposed Wasserstein GAN (WGAN) makes progress toward stable training of GANs, but sometimes can still generate only low-quality samples or fail to converge. We find that these problems are often due to the use of weight clipping in WGAN to enforce a Lipschitz constraint on the critic, which can lead to undesired behavior. We propose an alternative to clipping weights: penalize the norm of gradient of the critic with respect to its input. Our proposed method performs better than standard WGAN and enables stable training of a wide variety of GAN architectures with almost no hyperparameter tuning, including 101-layer ResNets and language models over discrete data. We also achieve high quality generations on CIFAR-10 and LSUN bedrooms.

Ishaan Gulrajani, Kundan Kumar, Faruk Ahmed et al.

Natural image modeling is a landmark challenge of unsupervised learning. Variational Autoencoders (VAEs) learn a useful latent representation and model global structure well but have difficulty capturing small details. PixelCNN models details very well, but lacks a latent code and is difficult to scale for capturing large structures. We present PixelVAE, a VAE model with an autoregressive decoder based on PixelCNN. Our model requires very few expensive autoregressive layers compared to PixelCNN and learns latent codes that are more compressed than a standard VAE while still capturing most non-trivial structure. Finally, we extend our model to a hierarchy of latent variables at different scales. Our model achieves state-of-the-art performance on binarized MNIST, competitive performance on 64x64 ImageNet, and high-quality samples on the LSUN bedrooms dataset.

Michael Cogswell, Faruk Ahmed, Ross Girshick et al.

One major challenge in training Deep Neural Networks is preventing overfitting. Many techniques such as data augmentation and novel regularizers such as Dropout have been proposed to prevent overfitting without requiring a massive amount of training data. In this work, we propose a new regularizer called DeCov which leads to significantly reduced overfitting (as indicated by the difference between train and val performance), and better generalization. Our regularizer encourages diverse or non-redundant representations in Deep Neural Networks by minimizing the cross-covariance of hidden activations. This simple intuition has been explored in a number of past works but surprisingly has never been applied as a regularizer in supervised learning. Experiments across a range of datasets and network architectures show that this loss always reduces overfitting while almost always maintaining or increasing generalization performance and often improving performance over Dropout.

Faruk Ahmed, Daniel Tarlow, Dhruv Batra

When evaluating computer vision systems, we are often concerned with performance on a task-specific evaluation measure such as the Intersection-Over-Union score used in the PASCAL VOC image segmentation challenge. Ideally, our systems would be tuned specifically to these evaluation measures. However, despite much work on loss-aware structured prediction, top performing systems do not use these techniques. In this work, we seek to address this problem, incorporating loss-aware prediction in a manner that is amenable to the approaches taken by top performing systems. Our main idea is to simultaneously leverage two systems: a highly tuned pipeline system as is found on top of leaderboards, and a traditional CRF. We show how to combine high quality candidate solutions from the pipeline with the probabilistic approach of the CRF that is amenable to loss-aware prediction. The result is that we can use loss-aware prediction methodology to improve performance of the highly tuned pipeline system.