Automatic target recognition (ATR) is crucial for safety-critical tasks such as navigation and surveillance, particularly in demanding military scenarios with unfamiliar terrains, harsh environments, and unseen object categories. Conventional and open-world object detectors often fail in these contexts, lacking exposure to such novel conditions. Meanwhile, Large Vision-Language Models (LVLMs) exhibit zero-shot recognition capabilities across diverse settings yet struggle with precise localization. We address these limitations by combining the localization strength of open-world detectors with the recognition confidence of LVLMs, creating a robust pipeline for zero-shot ATR in novel domains and classes. Our study compares several LVLMs on underrepresented military vehicles, examining factors such as distance range, modality, and prompting strategies. These findings provide insights for developing more reliable ATR systems in uncharted environments.

Implicit Neural Representations (INRs) are increasingly recognized as a versatile data modality for representing discretized signals, offering benefits such as infinite query resolution and reduced storage requirements. Existing signal compression approaches for INRs typically employ one of two strategies: 1. direct quantization with entropy coding of the trained INR; 2. deriving a latent code on top of the INR through a learnable transformation. Thus, their performance is heavily dependent on the quantization and entropy coding schemes employed. In this paper, we introduce SINR, an innovative compression algorithm that leverages the patterns in the vector spaces formed by weights of INRs. We compress these vector spaces using a high-dimensional sparse code within a dictionary. Further analysis reveals that the atoms of the dictionary used to generate the sparse code do not need to be learned or transmitted to successfully recover the INR weights. We demonstrate that the proposed approach can be integrated with any existing INR-based signal compression technique. Our results indicate that SINR achieves substantial reductions in storage requirements for INRs across various configurations, outperforming conventional INR-based compression baselines. Furthermore, SINR maintains high-quality decoding across diverse data modalities, including images, occupancy fields, and Neural Radiance Fields.

3D object detection and pose estimation from a single-view image is challenging due to the high uncertainty caused by the absence of 3D perception. As a solution, recent monocular 3D detection methods leverage additional modalities, such as stereo image pairs and LiDAR point clouds, to enhance image features at the expense of additional annotation costs. We propose using diffusion models to learn effective representations for monoc-ular 3D detection without additional modalities or training data. We present MonoDiff, a novel framework that em-ploys the reverse diffusion process to estimate 3D bounding box and orientation. But, considering the variability in bounding box sizes along different dimensions, it is inef-fective to sample noise from a standard Gaussian distribution. Hence, we adopt a Gaussian mixture model to sam-ple noise during the forward diffusion process and initialize the reverse diffusion process. Furthermore, since the diffusion model generates the 3D parameters for a given object image, we leverage 2D detection information to pro-vide additional supervision by maintaining the correspon-dence between 3D/2D projection. Finally, depending on the signal-to-noise ratio, we incorporate a dynamic weighting scheme to account for the level of uncertainty in the supervision by projection at different timesteps. MonoDiff outperforms current state-of-the-art monocular 3D detection methods on the KITTI and Waymo benchmarks without additional depth priors. MonoDiff project is available at: https://dylran.github.iolmonodiffgithub.io.

Crowd counting is a fundamental problem in crowd analysis which is typically accomplished by estimating a crowd density map and summing over the density values. However, this approach suffers from background noise accumulation and loss of density due to the use of broad Gaussian kernels to create the ground truth density maps. This issue can be overcome by narrowing the Gaussian kernel. However, existing approaches perform poorly when trained with ground truth density maps with broad kernels. To deal with this limitation, we propose using conditional diffusion models to predict density maps, as diffusion models show high fidelity to training data during generation. With that, we present CrowdDiff that generates the crowd density map as a reverse diffusion process. Further-more, as the intermediate time steps of the diffusion process are noisy, we incorporate a regression branch for direct crowd estimation only during training to improve the feature learning. In addition, owing to the stochastic nature of the diffusion model, we introduce producing multiple density maps to improve the counting performance contrary to the existing crowd counting pipelines. We conduct extensive experiments on publicly available datasets to validate the effectiveness of our method. CrowdDiff out-performs existing state-of-the-art crowd counting methods on several public crowd analysis benchmarks with significant improvements. CrowdDiff project is available at: https://dylran.github.io/crowddiffgithub.io.

Crowd localization is a fundamental problem in understanding a crowded scene as it provides salient information such as location, density, and count about the crowd. However, the state-of-the-art face detection methods incorporate bounding box annotations, face landmark coordinates, and other structural ground truth information, which is expensive to prepare compared to the point annotation of faces. On the other hand, the state-of-the-art crowd localization methods are based on regressing the point coordinates and classification scores, which hinders the ability to learn the distributions of point coordinates and confidence scores for a given crowd dataset. The learned distribution allows us to infer new insights for a given crowd image, which in turn helps improve the confidence scores for positive sample detection. We propose solving the crowd localization task as a generative method using recently proposed denoising diffusion probabilistic models. Furthermore, we leverage the stochastic nature of the generative model by using multiple realizations of localization results for a given crowd image to improve crowd detection performance. We conduct extensive experiments on publicly available datasets to validate the effectiveness of our method. The proposed method improves localization performance metrics: precision, recall, and F1-measure by at least 2% for UCF-QNRF and ShanghaiTech A datasets and achieves state-of-the-art counting performance across five public crowd counting datasets.