{"id":"3dad84aa-71e3-4a92-9726-43ba5c6e2785","arxiv_id":"2605.12566","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":3.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"A Swin Transformer-based semantic communication framework with federated learning reports 5.7 dB PSNR gains over DeepJSCC baselines for UAV image transmission on CIFAR-10.","lead":"The paper proposes a Swin Transformer semantic communication system paired with federated learning to transmit images from UAVs while meeting bandwidth limits and privacy rules. A smart generalist might read it to see how AI can make drone image sharing more practical in real-world low-altitude networks.","discovery_kind":"new_application","skeptic_critique":{"model":"grok-4.3","headline":"Simulation-only PSNR gains on CIFAR-10 rest on unverified assumptions about channel modeling and UAV dynamics.","rationale":"The reader's weakest assumption correctly flags the sim-to-real gap. Full-text methods would need to show explicit UAV channel realism for the claim to hold; absent that, the performance numbers remain provisional. This check directly tests transferability without requiring new hardware.","tokens_in":1740,"tokens_out":337,"duration_ms":21754,"concrete_test":"Re-implement the STSC encoder/decoder and FL training loop; replace the original channel with a Rician fading model parameterized by UAV altitude 50-150 m and velocity 5-20 m/s (as typical in low-altitude scenarios); recompute PSNR on the same CIFAR-10 test set at the reported bandwidth ratios. If the advantage over DeepJSCC falls below 3 dB or convergence slows, the central claim does not transfer.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The headline claim of ≥5.7 dB PSNR improvement over DeepJSCC baselines is derived from simulations on CIFAR-10. For this to support the STSC+FL framework in low-altitude networks, the experimental channel must faithfully capture UAV-specific impairments (variable path loss, Doppler, interference) and bandwidth constraints. If the methods section uses only static AWGN or simplified models without these, the reported gain and convergence advantages become artifacts of the simulation rather than evidence of real deployment viability. The FL privacy mechanism adds another untested layer: any utility loss from federated aggregation is not quantified against the PSNR metric.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript proposes a Swin Transformer-based Semantic Communication (STSC) framework integrated with Federated Learning (FL) for privacy-preserving image transmission in low-altitude UAV networks. It designs multi-scale semantic feature extraction under bandwidth constraints, deploys dedicated nodes on UAVs, and uses FL to train without sharing raw data. Simulations on CIFAR-10 are reported to yield at least 5.7 dB PSNR improvement over DeepJSCC baselines together with better convergence and generalization.","tokens_in":1856,"tokens_out":489,"duration_ms":30162,"significance":"If the performance claims are robustly supported, the work could advance semantic communications for bandwidth-limited, privacy-sensitive UAV applications by combining transformer-based feature extraction with distributed training. The integration of SC and FL addresses a timely problem in low-altitude networks, though the simulation-only evidence on a standard image dataset restricts immediate claims about real-world UAV viability.","major_comments":[{"comment":"Simulation Experiments section: the headline claim of ≥5.7 dB PSNR gain over DeepJSCC is presented without an experimental protocol, baseline implementation details, error bars, or statistical tests, leaving the central performance result weakly supported.","section":"Simulation Experiments"},{"comment":"Methods / Channel Model subsection: the simulations appear to rely on static or simplified channel models (e.g., AWGN) without explicit incorporation of UAV-specific impairments such as variable path loss, Doppler shifts, or interference; this undermines applicability to the stated low-altitude network setting.","section":"Methods"}],"minor_comments":[{"comment":"Abstract: the phrase 'at least 5.7 dB' should be accompanied by the precise SNR, bandwidth, and model-size conditions under which the gain is measured.","section":"Abstract"},{"comment":"Notation: the definition of semantic feature maps and the FL aggregation rule should be stated explicitly with equation numbers to avoid ambiguity when comparing to DeepJSCC.","section":"STSC Architecture"}],"recommendation":"major_revision","confidential_remarks":"The manuscript's reliance on CIFAR-10 and simplified channels is a common limitation in this area; the editor may wish to request additional ablation studies with realistic UAV channel traces if the authors can supply them within the revision scope."},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive comments. We address each major comment point by point below, indicating the revisions we will make to improve the manuscript.","responses":[{"response":"We agree that additional details are required to robustly support the central performance claim. In the revised manuscript, we will expand the Simulation Experiments section with a complete experimental protocol (including hyperparameters, training schedules, and data splits), explicit implementation details for the DeepJSCC baselines, results reported with error bars from multiple independent runs, and statistical significance tests (e.g., paired t-tests) to validate the reported PSNR gains.","revision_made":"yes","referee_comment":"[Simulation Experiments] Simulation Experiments section: the headline claim of ≥5.7 dB PSNR gain over DeepJSCC is presented without an experimental protocol, baseline implementation details, error bars, or statistical tests, leaving the central performance result weakly supported."},{"response":"The current work uses an AWGN model as a controlled baseline to isolate the contributions of the Swin Transformer semantic extractor and federated learning under bandwidth limits. We acknowledge that this simplification limits direct applicability to real low-altitude UAV channels. In the revision, we will augment the Channel Model subsection with a discussion of UAV-specific impairments and include additional simulation results that incorporate standard models for path loss and Doppler shift. Full modeling of dynamic interference and hardware effects is noted as future work, as it would require specialized UAV channel datasets beyond the scope of this study.","revision_made":"partial","referee_comment":"[Methods] Methods / Channel Model subsection: the simulations appear to rely on static or simplified channel models (e.g., AWGN) without explicit incorporation of UAV-specific impairments such as variable path loss, Doppler shifts, or interference; this undermines applicability to the stated low-altitude network setting."}],"tokens_in":1338,"tokens_out":402,"duration_ms":40998,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main takeaway is that the authors combine a Swin Transformer for multi-scale semantic feature extraction with semantic communication and federated learning to handle bandwidth limits and privacy in low-altitude UAV image transmission. They report at least 5.7 dB better PSNR than DeepJSCC baselines from CIFAR-10 simulations, plus improved convergence and generalization. The framework also includes dedicated comm and compute nodes on the UAVs for flexibility in logistics or inspection tasks. This integration is a reasonable engineering step for the specific setting, as Swin handles hierarchical features well under constraints and FL avoids raw data sharing. The architecture choices line up with the stated goals of efficiency and privacy. The soft spots sit in the experiments. Everything rests on simulations with no reported details on the channel model, so it is unclear whether variable path loss, Doppler, or interference typical of UAV flights were included. CIFAR-10 is a convenient dataset but does not match aerial imagery, and the paper supplies no error bars, statistical tests, or separate quantification of any utility loss from federated aggregation. Without those elements the reported gains are hard to interpret as evidence of deployment viability rather than simulation artifacts. This paper is for engineers and researchers working on practical UAV communication systems or semantic comm applications. A reader already familiar with the base techniques could pick up implementation ideas for the combined setup, though they would need to run their own tests to confirm the numbers. The work shows clear engagement with the relevant components and problem constraints, so it deserves peer review to examine the methods section and push for stronger validation.","headline":"This applies Swin Transformer semantic comm and federated learning to UAV image transmission and claims a 5.7 dB PSNR gain on CIFAR-10 simulations, but the evaluation leaves real-world UAV channel effects untested.","tokens_in":2387,"tokens_out":403,"would_cite":false,"duration_ms":48118,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":{"model":"grok-4.3","evidence":[],"headline":"Paper's STSC+FL UAV image transmission framework has no structural overlap with RS forcing chain or J-cost machinery","alignment":"orthogonal","rationale":"The paper's central machinery (Swin Transformer hierarchical encoder-decoder with patch-merging, FedAvg aggregation, MSE loss on CIFAR-10, AWGN/Rician/Rayleigh channel layers) is standard empirical ML/comms engineering. It contains no J-cost functions, golden-ratio ladders, 8-tick periodicity, ratio-symmetric costs, or parameter-free constant derivations. RS theorems such as reality_from_one_distinction, Jcost uniqueness via Aczél, and AlexanderDuality D=3 forcing have no counterpart here; the work operates entirely in the applied domain of semantic JSCC and cross-silo FL, where RS supplies neither predictions nor contradictions.","tokens_in":55300,"confidence":"high","tokens_out":187,"duration_ms":9545,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"A Swin Transformer semantic communication system with federated learning improves UAV image transmission quality by at least 5.7 dB PSNR while keeping raw data private.","keywords":["semantic communication","federated learning","Swin Transformer","UAV image transmission","privacy preservation","low-altitude networks","PSNR improvement","bandwidth-constrained transmission"],"falsifier":"A field experiment with actual UAVs sending real images over live wireless links that shows no PSNR gain or leaks raw data would disprove the central performance and privacy claims.","tokens_in":2638,"feed_emoji":"📡","tokens_out":632,"duration_ms":26496,"temperature":0.7,"pith_summary":"The paper proposes a semantic communication framework called STSC for transmitting images from UAVs to ground stations under tight bandwidth constraints. It uses a Swin Transformer to extract multi-scale semantic features from images and combines this with federated learning so that models train across devices without any raw image data leaving the UAVs. Simulations on the CIFAR-10 dataset show the approach delivers higher reconstructed image quality than standard DeepJSCC methods and converges more reliably. The work targets practical low-altitude applications such as logistics and inspection where both bandwidth and privacy rules are strict.","feed_headline":"Transformer boosts UAV image PSNR by 5.7 dB with privacy","feed_subtitle":"Semantic communication and federated learning let drones send clearer images without sharing raw data under tight bandwidth.","key_machinery":"The Swin Transformer-based Semantic Communication (STSC) architecture, which extracts multi-scale semantic features from images for bandwidth-efficient transmission and integrates federated learning to train models across UAVs without sharing raw data.","core_discovery":"The STSC architecture extracts multi-scale semantic features via a Swin Transformer under bandwidth limits, pairs it with federated learning for distributed training without raw data exchange, and achieves at least 5.7 dB higher PSNR on CIFAR-10 reconstructions than DeepJSCC baselines while improving convergence and generalization.","pith_inferences":["The same architecture could be adapted to transmit other sensor streams such as video or LiDAR from UAVs.","If the privacy mechanism scales, it might support multi-UAV swarms sharing semantic updates without a central server.","Variable real-world channel fading not present in CIFAR-10 simulations remains an open variable for deployment."],"forward_implications":["UAV image transmissions maintain higher visual quality despite severe bandwidth limits.","Raw images never leave the UAV, satisfying strict privacy rules for distributed operations.","Dedicated on-board nodes allow flexible real-time coverage without central data aggregation.","The model shows faster convergence and better generalization across different transmission scenarios."],"fun_headline_variants":["Swin Transformer FL boosts UAV PSNR by 5.7 dB","STSC Swin model gains 5.7 dB PSNR for UAV privacy","FL and Swin Transformer raise UAV PSNR 5.7 dB privately","STSC framework with FL improves PSNR by 5.7 dB in UAVs"],"cache_read_input_tokens":64,"weakest_assumption_plain":"That performance gains measured on CIFAR-10 under simulated conditions will hold for real UAV deployments facing actual bandwidth limits, channel noise, and privacy regulations.","fun_headline_variants_meta":{"raw":{"variants":["Swin Transformer FL boosts UAV PSNR by 5.7 dB","STSC Swin model gains 5.7 dB PSNR for UAV privacy","FL and Swin Transformer raise UAV PSNR 5.7 dB privately","STSC framework with FL improves PSNR by 5.7 dB in UAVs"]},"model":"grok-4.3","cost_usd":0.012309,"raw_usage":{"total_tokens":5278,"prompt_tokens":655,"num_sources_used":0,"completion_tokens":87,"cost_in_usd_ticks":123090500,"prompt_tokens_details":{"text_tokens":655,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":4536,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":655,"tokens_out":87,"duration_ms":59655,"temperature":1.0,"reasoning_tokens":4536,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-05-14T20:45:25.535252+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A field experiment with actual UAVs sending real images over live wireless links that shows no PSNR gain or leaks raw data would disprove the central performance and privacy claims.","supporting_citations":[],"review_version":1}