Noor Ahmed

Tom Fischer, Yaoyao Liu, Artur Jesslen et al.

Different from human nature, it is still common practice today for vision tasks to train deep learning models only initially and on fixed datasets. A variety of approaches have recently addressed handling continual data streams. However, extending these methods to manage out-of-distribution (OOD) scenarios has not effectively been investigated. On the other hand, it has recently been shown that non-continual neural mesh models exhibit strong performance in generalizing to such OOD scenarios. To leverage this decisive property in a continual learning setting, we propose incremental neural mesh models that can be extended with new meshes over time. In addition, we present a latent space initialization strategy that enables us to allocate feature space for future unseen classes in advance and a positional regularization term that forces the features of the different classes to consistently stay in respective latent space regions. We demonstrate the effectiveness of our method through extensive experiments on the Pascal3D and ObjectNet3D datasets and show that our approach outperforms the baselines for classification by $2-6\%$ in the in-domain and by $6-50\%$ in the OOD setting. Our work also presents the first incremental learning approach for pose estimation. Our code and model can be found at https://github.com/Fischer-Tom/iNeMo.

Noor Ahmed, Anna Kukleva, Bernt Schiele

Few-Shot Class-Incremental Learning (FSCIL) introduces a paradigm in which the problem space expands with limited data. FSCIL methods inherently face the challenge of catastrophic forgetting as data arrives incrementally, making models susceptible to overwriting previously acquired knowledge. Moreover, given the scarcity of labeled samples available at any given time, models may be prone to overfitting and find it challenging to strike a balance between extensive pretraining and the limited incremental data. To address these challenges, we propose the OrCo framework built on two core principles: features' orthogonality in the representation space, and contrastive learning. In particular, we improve the generalization of the embedding space by employing a combination of supervised and self-supervised contrastive losses during the pretraining phase. Additionally, we introduce OrCo loss to address challenges arising from data limitations during incremental sessions. Through feature space perturbations and orthogonality between classes, the OrCo loss maximizes margins and reserves space for the following incremental data. This, in turn, ensures the accommodation of incoming classes in the feature space without compromising previously acquired knowledge. Our experimental results showcase state-of-the-art performance across three benchmark datasets, including mini-ImageNet, CIFAR100, and CUB datasets. Code is available at https://github.com/noorahmedds/OrCo

Noor Ahmed, Cameron Braunstein, Steffen Eger et al.

Recent Multi-Modal Large Language Models (MLLMs) have demonstrated strong capabilities in learning joint representations from text and images. However, their spatial reasoning remains limited. We introduce 3DFroMLLM, a novel framework that enables the generation of 3D object prototypes directly from MLLMs, including geometry and part labels. Our pipeline is agentic, comprising a designer, coder, and visual inspector operating in a refinement loop. Notably, our approach requires no additional training data or detailed user instructions. Building on prior work in 2D generation, we demonstrate that rendered images produced by our framework can be effectively used for image classification pretraining tasks and outperforms previous methods by 15%. As a compelling real-world use case, we show that the generated prototypes can be leveraged to improve fine-grained vision-language models by using the rendered, part-labeled prototypes to fine-tune CLIP for part segmentation and achieving a 55% accuracy improvement without relying on any additional human-labeled data.

Mubark Jedh, Lotfi ben Othmane, Noor Ahmed et al.

The smart features of modern cars are enabled by a number of Electronic Control Units (ECUs) components that communicate through an in-vehicle network, known as Controller Area Network (CAN) bus. The fundamental challenge is the security of the communication link where an attacker can inject messages (e.g., increase the speed) that may impact the safety of the driver. Developing an effective defensive security solution depends on the knowledge of the identity of the ECUs, which is proprietary information. This paper proposes a message injection attack detection mechanism that is independent of the IDs of the ECUs, which is achieved by capturing the patterns in the message sequences. First, we represent the sequencing ofther messages in a given time-interval as a direct graph and compute the similarities of the successive graphs using the cosine similarity and Pearson correlation. Then, we apply threshold, change point detection, and Long Short-Term Memory (LSTM)-Recurrent NeuralNetwork (RNN) to detect and predict malicious message injections into the CAN bus. The evaluation of the methods using a dataset collected from a moving vehicle under malicious RPM and speed reading message injections show a detection accuracy of 98.45% when using LSTM-RNN and 97.32% when using a threshold method. Further, the pace of detecting the change isfast for the case of injection of RPM reading messagesbut slow for the case of injection of speed readingsmessages.