--- title: 'GeoSapiens: Few-Shot Dental Landmark Detection' url: https://www.emergentmind.com/topics/geosapiens type: topic --- # GeoSapiens: Few-Shot Dental Landmark Detection Searching arXiv for the specified paper and closely related references to ground the article in current research. GeoSapiens is a few-shot learning framework for dental landmark detection on cone-beam computed tomography (CBCT) of anterior teeth. It was introduced in "Geometric-Guided Few-Shot Dental Landmark Detection with Human-Centric Foundation Model" [2507.04710] to address a setting in which manual landmark annotation by dentists is time-consuming, labor-intensive, and subject to inter-observer variability, while conventional deep learning is constrained by scarce training data and the high cost of expert annotations. The framework combines a baseline adapted from Sapiens, a human-centric foundation model, with a geometric loss designed to encode anatomical structure, and it is reported to surpass existing landmark detection methods on the authors' collected dataset of anterior teeth landmarks, including an 8.18% higher success detection rate at a strict 0.5 mm threshold than the leading approach [2507.04710]. ## 1. Clinical task and problem formulation Accurate detection of anatomic landmarks is described as essential for assessing alveolar bone and root conditions, thereby optimizing clinical outcomes in orthodontics, periodontics, and implant dentistry [2507.04710]. Within this setting, GeoSapiens targets anterior-teeth CBCT slices and predicts 16 distinct anatomical landmarks per slice on upper and lower anterior teeth. The dataset used in the study, LDTeeth, contains 347 images of size \(957 \times 555\) px and 5,552 total landmark points, with 16 landmarks for each image [2507.04710]. The paper frames the task as few-shot adaptation rather than episodic meta-learning. The training set consists of only 36 CBCT slices drawn from 3 patients, with validation on 149 images from 14 patients and testing on 162 images from 16 patients [2507.04710]. This design emphasizes fine-tuning under severe annotation scarcity. A plausible implication is that the method is intended less as a generic meta-learning algorithm than as a parameter-efficient transfer pipeline for small clinical datasets. The operating unit is a 2D single mid-sagittal slice. Each slice is first cropped to a dentist-annotated tooth bounding box corresponding to the full mid-sagittal cut, after which the model predicts landmark heat maps and coordinates [2507.04710]. The paper explicitly notes that out-of-plane lesions are not addressed, which bounds the current formulation to 2D localization rather than full volumetric landmarking [2507.04710]. ## 2. Architectural design GeoSapiens uses Sapiens-0.3B as its backbone, a Vision Transformer foundation model pre-trained via Masked Autoencoding on 300 M in-the-wild human images [2507.04710]. The motivation given in the paper is a transfer hypothesis: anterior teeth CBCT slices share symmetric, morphological patterns akin to human body parts, so Sapiens representations may transfer effectively [2507.04710]. This is presented as a hypothesis rather than a formally proven property. Instead of full fine-tuning, the framework inserts Low-Rank Adapters (LoRA) into all multi-head self-attention projections and multilayer perceptrons. The attention \(Q/K/V\) projections use rank \(r=4\) with \(\alpha=4\), and the feed-forward projection layers use \(r=8\) with \(\alpha=8\) [2507.04710]. Full fine-tuning would involve approximately 330 M trainable weights, whereas the LoRA configuration reduces trainable parameters to approximately 24 M, reported as approximately 0.3% of the original in the summary, while the implementation details also describe the reduction as \(330\text{ M} \to 24\text{ M}\) [2507.04710]. A top-down heat-map head, described as following SimpleBaseline, transforms ViT features into \(K=16\) per-landmark likelihood maps \(H_k(x,y)\) [2507.04710]. During training, the model uses differentiable soft-argmax to obtain continuous 2D coordinates: $$ \hat p_k = \sum_{x,y} (x,y)\;M_k(x,y)\,,\quad M_k(x,y)=\frac{\exp\bigl(H_k(x,y)/T\bigr)}{\sum_{u,v}\exp\bigl(H_k(u,v)/T\bigr)}\,,\;T=0.1\,. $$ At inference time, arg-max on the heat map recovers integer coordinates [2507.04710]. The complete pipeline is summarized in the paper as CBCT \(\to\) ViT+LoRA \(\to\) heat-map head \(\to\) soft-argmax \(\to\) coordinate and geometric loss [2507.04710]. In functional terms, GeoSapiens is therefore a heat-map regression model with parameter-efficient foundation-model adaptation and explicit geometric regularization. ## 3. Geometric priors and loss construction A defining feature of GeoSapiens is its geometric loss, which encodes two stated priors on dental landmark configurations. First, the tooth-axis line (AP–CP) is perpendicular to three level lines: root apex, apical-third, and mid-root. Second, these three level lines are mutually parallel [2507.04710]. These priors are imposed after extracting the 16 landmark points. The method fits each relevant line by least squares and computes unit direction vectors: - \(\mathbf{v}_\perp\) for the AP–CP axis, - \(\mathbf{v}_1,\mathbf{v}_2,\mathbf{v}_3\) for the three horizontal levels [2507.04710]. The geometric term is then defined as $$ \mathcal{L}_{\rm geo} = \frac{1}{6}\Bigl( \sum_{j=1}^{3}\mathbf{v}_\perp\!\cdot\!\mathbf{v}_j \;+\; \sum_{1\le j