{"total":12,"items":[{"citing_arxiv_id":"2606.18816","ref_index":14,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"SwitchBraidNet: Quantisation-Aware Lightweight Architecture for Hybrid Brain-Computer Interface","primary_cat":"cs.HC","submitted_at":"2026-06-17T08:35:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"SwitchBraidNet is a compact dual-path EEG classifier achieving 69.49% MI accuracy (FP16), 93.48% SSVEP accuracy (FP32), 64.82 bits/min hybrid ITR (FP16), and 3.03 KB INT8 size via quantization-aware training on OpenBMI.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.31329","ref_index":16,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Improving acoustic drone detection generalization through pretraining and data augmentation","primary_cat":"eess.AS","submitted_at":"2026-05-29T14:04:55+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Pretraining on broad sound events plus on-the-fly augmentations improves out-of-domain true-positive rates for acoustic drone detection at fixed low false-positive rates.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.11989","ref_index":28,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"A Transfer Learning Evaluation of Deep Neural Networks for Image Classification","primary_cat":"cs.CV","submitted_at":"2026-05-12T11:40:49+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":2.0,"formal_verification":"none","one_line_summary":"Empirical comparison of transfer learning performance across eleven pre-trained models on five image datasets using accuracy, time, and size metrics.","context_count":1,"top_context_role":"background","top_context_polarity":"support","context_text":"by [27] by skipping some intermediate layers to allow the special flow of information across the layers, for example zero-padding, projection, dropout, skip connections, etc; • Channels: CNNs have powerful performance in learning features automatically, and this can be dynamically performed by tuning the kernel weights. However, some feature maps have little or no role in object discrimination [28] and could cause overfitting as well. Those feature maps (or the channels) can be optimally selected in designing the CNN to avoid overfitting. 3.2. Neural Network Architectures This study tested eleven popular pre-trained models. Figure 1 gives a comprehensive infographic representation over time. Table 1 depicts all the tested models with their main"},{"citing_arxiv_id":"2605.07828","ref_index":12,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"NSPOD: Accelerating Krylov solvers via DeepONet-learned POD subspaces","primary_cat":"math.NA","submitted_at":"2026-05-08T14:56:46+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"NSPOD is a multigrid-like preconditioner using DeepONet-learned POD subspaces that dramatically cuts Krylov solver iterations for solid mechanics PDEs on unstructured CAD geometries, outperforming algebraic multigrid.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":"stacked layers of PointTransformer blocks [48]. PointTransformers are powerful architectures that adapt traditional transformerstopointclouddata;theyareabletocapturelong-rangedependenciesbetweenpointsin3Dspace.Finally, attheendofthebranchsubnetworkwealsoincludeacustomizedsqueezed-and-excitationnetwork,producingascaled outputactingasanattentionmechanism[12,20,42].BycombiningallthesecomponentswebuildPTFONet.Wetrain thesePTFONetsonanumberofCADgeometries,severalofthemifwewanttoachieveacertaindegreeofgeometric transferability. We then use these trained networks to perform solution inference and build preconditioners based on Proper Orthogonal Decomposition (POD) of these inferred solutions; we denote this hybrid preconditioner as Neural"},{"citing_arxiv_id":"2605.01484","ref_index":16,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Evaluating LLMs on Large-Scale Graph Property Estimation via Random Walks","primary_cat":"cs.LG","submitted_at":"2026-05-02T15:11:52+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"EstGraph benchmark evaluates LLMs on estimating properties of very large graphs from random-walk samples that fit in context limits.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.01283","ref_index":148,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Developing a Strong Pre-Trained Base Model for Plant Leaf Disease Classification","primary_cat":"cs.CV","submitted_at":"2026-05-02T06:33:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"A DenseNet201 base model trained on a constructed plant leaf disease dataset outperforms baselines and enables faster, more robust transfer learning with less data than general models.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.16304","ref_index":31,"ref_count":2,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Modulation Feature Enhancement with a Multi-Stage Attention Network for Underwater Acoustic Target Recognition","primary_cat":"eess.SP","submitted_at":"2026-04-24T18:10:57+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A 1-D CNN with novel multi-stage spectral attention mechanisms and adjustable class-balanced focal loss improves recognition accuracy on real ship-radiated noise datasets.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.17688","ref_index":27,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Dual-stream Spatio-Temporal GCN-Transformer Network for 3D Human Pose Estimation","primary_cat":"cs.CV","submitted_at":"2026-04-20T01:07:44+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"MixTGFormer reports state-of-the-art 3D pose estimation errors of 37.6 mm on Human3.6M and 15.7 mm on MPI-INF-3DHP by using parallel GCN-Transformer streams with SE layers for local-global feature fusion.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"The MixTGFormer, built on Mixformer, outperforms other state -of-the-art methods on the Human3.6M and MPI-INF-3DHP datasets, achieving the best performance. 2. Related Work 2.1. 3D Human Pose Estimation 3D human pose estimation is a classic and important problem in the field of computer vision, with decades of research history [26]. In the early stages [27,28,29], this work relied almost entirely on handcrafted features and geometric constraints as means to predict 3D human poses. With the rapid development of deep learning, deep learning has now become the primary method for 3D human pose estimation [30]. This problem can be classified from different perspectives, such as based on input data and estimation methods."},{"citing_arxiv_id":"2603.07119","ref_index":27,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TIQA: Human-Aligned Perceptual Text Quality Assessment in Generated Images","primary_cat":"cs.CV","submitted_at":"2026-03-07T09:11:10+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"TIQA introduces datasets and a model that predict human perceptual quality of rendered text in AI images, achieving PLCC 0.942 on crops and improving selected image text quality by 0.36 MOS.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2602.13636","ref_index":36,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"Layer-Guided UAV Tracking: Enhancing Efficiency and Occlusion Robustness","primary_cat":"cs.CV","submitted_at":"2026-02-14T07:02:25+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"LGTrack achieves 258.7 FPS real-time UAV tracking with 82.8% precision on UAVDT by combining dynamic layer selection, Global-Grouped Coordinate Attention, and Similarity-Guided Layer Adaptation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2504.13102","ref_index":14,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"A Multi-task Learning Balanced Attention Convolutional Neural Network Model for Few-shot Underwater Acoustic Target Recognition","primary_cat":"cs.SD","submitted_at":"2025-04-17T17:11:32+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"MT-BCA-CNN achieves 97% accuracy and 95% F1-score on 27-class few-shot underwater acoustic target recognition by combining channel attention and multi-task learning on the Watkins Marine Life Dataset.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2403.16958","ref_index":25,"ref_count":1,"confidence":0.9,"is_internal_anchor":false,"paper_title":"TwinLiteNet+: An Enhanced Multi-Task Segmentation Model for Autonomous Driving","primary_cat":"cs.CV","submitted_at":"2024-03-25T17:17:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"TwinLiteNet+ is a hybrid-encoder multi-task segmentation model with new UCB, USB, and PCAA modules that reports 92.9% mIoU on drivable area and 34.2% IoU on lane segmentation on BDD100K while using 11x fewer FLOPs than prior models.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}