Library · Signals, sensing & imaging
What Semantic & Object Recognition got wrong, from 22 theses
Summary written by a model from the records below: The collected records detail practical bottlenecks in semantic and object recognition across aerial imagery, robotics, and medical applications. Investigators encountered severe computational constraints, boundary confusion in overlapping objects, labeling bottlenecks, and distribution shifts between capture domains. These records come from graduate theses, 2021 to 2026, each labeled with its degree where the repository states it, and each links to its thesis. They were extracted by language models reading the full text, so treat each as a lead to read, not a verdict.
Plus 10 problems the authors left open, listed at the end.
Failure modes, by theses reporting them
Select a mode to jump to its records.
Excessive computational overhead and latency in semantic segmentation models
Pixel-level semantic segmentation and complex fusion or transformer architectures impose prohibitive computational expense and latency on embedded hardware and high-resolution aerial imagery. In several robotics applications, these models achieved unacceptably low frame rates or proved unnecessary compared to simpler bounding-box centroids and manual methods.
Considered and rejected
Considered and rejected: Semantic segmentation was dropped in favor of object detection bounding-box centroids for downstream goal generation due to the computational overhead and latency of multi-channel segmentation fusion models on embedded hardware.
Autonomous System for Identifying and Capturing Floating Waste · Georgia Tech · PhD
Lost to a baseline
VGG-16 semantic segmentation achieved only 2 fps with frequent misidentifications and lost to manual contact determination
Characterisation of a Novel Bioadhesive Used by the Ctenophore Pleurobrachia pileus · University College Dublin · PhD
Considered and rejected
Considered and rejected: Rejected pixel-wise semantic/instance segmentation and GrabCut for object extraction due to high computational overhead in robotics.
Scene understanding via scene graph Szenenverständnis mittels Szenengraph · Leibniz Universität Hannover · PhD
Considered and rejected
Considered and rejected: Pixel-level semantic segmentation of aerial imagery, rejected due to extreme computational overhead on ultra-high-resolution images and variable scale/resolution distortions across oblique aerial perspectives.
ASSESSING THE QUALITY OF PREDICTIVE UNCERTAINTY ON COASTAL AERIAL IMAGES WITH BAYESIAN TRANSFER LEARNING · Naval Postgraduate School · Master's
Considered and rejected
Considered and rejected: Decided against semantic segmentation architectures (pixel-wise segmentation) due to high computational cost and because area composition estimation does not require precise item boundary segmentation.
Development of a method to classify and analyse the composition of mixed waste materials in real-time. · Cranfield · PhD
Considered and rejected
Considered and rejected: Decided against using the fusion CNN+GNN architecture for semantic vessel segmentation due to grid-graph generation overhead and class fragmentation
Application of machine learning methods to the analysis of x-ray angiography images · Oxford · PhD
Considered and rejected
Considered and rejected: Decided against transformer architectures for semantic segmentation due to extreme computational expense and large data requirements compared to CNNs.
Integration of machine learning for enhanced digital rock physics workflows · UT Austin · PhD
Boundary confusion and class merging for overlapping or cluttered objects
Semantic segmentation methods struggle when distinct localized patterns or obstacles appear in close spatial proximity, frequently merging discrete entities into a single continuous class. Coarse localization, reliance on global shape cues, and dense visual clutter cause boundary classification errors and hyperbola overlap issues.
Tried and failed
semantic segmentation of spatially adjacent overlapping features applied to medical image segmentation of multiple findings. Outcome: did not generalise. Reason: models confuse boundary distinctions when multiple distinct localized patterns co-occur in close spatial proximity
AI Systems for Understanding and Grounding Radiology Reports · Harvard · Master's
Tried and failed
Fully convolutional networks using global shape cues applied to satellite semantic segmentation. Outcome: no signal. Reason: Global shape cues failed to provide sufficient discriminative features for accurate boundary classification
Mapping agricultural land in support of opium monitoring in Afghanistan with Convolutional Neural Networks (CNNs). · Cranfield · PhD
Tried and failed
semantic segmentation without instance segmentation for obstacle mapping applied to robot path planning in outdoor environments. Reason: distinct obstacles were merged into a single continuous class, blocking navigable paths
Perception Enabled Planning for Autonomous Systems · Cornell · PhD
Considered and rejected
Considered and rejected: Rejected block-based GPR classification in favor of pixel-level semantic segmentation due to coarse localization and inability to handle overlapping hyperbolas
Surface and Sub-Surface Analyses for Bridge Inspection · University of Nevada, Reno · PhD
Performance degradation from deeper backbones, rigid cropping, or suboptimal sampling
Deeper convolutional backbones and unadapted real-time one-stage models degraded accuracy across segmentation benchmarks or showed severe sensitivity to input resolution. Arbitrary bounding box crop margins truncated target boundaries while random background sampling underperformed targeted feature-sampling schemes.
Tried and failed
deeper convolutional backbone in encoder-decoder network applied to semantic image segmentation. Outcome: worse than baseline. Reason: deeper architecture degraded performance across all evaluation metrics compared to shallower baseline
3D Panoptic Segmentation with Unsupervised Clustering for Visual Perception in Autonomous Driving · Cranfield · PhD
Tried and failed
fixed margin bounding box cropping and resizing applied to semantic image segmentation. Outcome: worse than baseline. Reason: fixed margin crops either truncate peripheral features or introduce irrelevant distracting background context
Deep face tracking and parsing in the wild · Imperial · PhD
Considered and rejected
Considered and rejected: Rejected DeepLabV3 and standard Unet architectures in favor of Unet++ for colon semantic segmentation due to superior IOU accuracy.
Fluorescence Imaging and Signal Quantification for Enhanced Surgical Precision and Real-Time Decision Making · University College Dublin · Doctorate
Tried and failed
real-time one-stage instance segmentation applied to sidewalk semantic segmentation. Reason: produced inaccurate segmentations and exhibited high sensitivity to input image resolution without specific backbone tuning
Robust and precise sidewalk detection with ensemble learning: Enhancing road safety and facilitating curb space management · Iowa State · Master's
Tried and failed
random background sampling for semantic segmentation applied to satellite imagery land cover classification. Outcome: worse than baseline. Reason: underperformed compared to boundary-targeted and feature-based sampling strategies
Mapping agricultural land in support of opium monitoring in Afghanistan with Convolutional Neural Networks (CNNs). · Cranfield · PhD
Generalization failure caused by domain shifts and unmodeled spatial context
Models trained under specific viewing angles or standard losses failed when deployed across camera domain shifts, such as transitioning from handheld images to aerial perspectives. Omitting topographic priors caused false positive predictions above elevation thresholds, while generic detectors failed to generalize to novel or unquantifiable objects.
Considered and rejected
Considered and rejected: Rejected using object recognition labels (YOLOv5s) directly for grasp selection, because household items are unquantifiable and novel objects fail.
Integrating computer vision with neuromuscular interfacing for the semi-autonomous control of robotic limbs · Imperial · PhD
Tried and failed
omitting digital elevation data in visual segmentation applied to aerial imagery semantic segmentation. Outcome: worse than baseline. Reason: lacks crucial topographic priors, causing false positive segmentations above natural elevation thresholds
Monitoring and understanding treeline dynamics in the Swiss Alps from 80 years of aerial imagery · EPFL · PhD
Tried and failed
recall loss objective applied to semantic image segmentation. Outcome: did not generalise. Reason: failed on highly complex scenes with intra-class similarities and saturated simple scenes
Robustness under distribution shifts in computer vision · Georgia Tech · PhD
Tried and failed
semantic segmentation models trained on handheld imagery applied to aerial bridge inspection imagery. Outcome: did not generalise. Reason: distribution shift between human-collected handheld images and unmanned aerial vehicle perspective and quality
Enhancing Structural Inspections by Integrating Computer Vision, Machine Learning, and Risk Assessment for Comprehensive Corrosion Evaluation · Georgia Tech · PhD
Prohibitive manual annotation overhead and small dataset instability
Pixel-level annotation demands prohibitive human labor costs compared to lightweight clustering techniques and leads to severe instability when training sets fall below five hundred images. Synthetic data augmentation for majority classes provided diminishing returns when sufficient real features already existed.
Tried and failed
U-Net semantic segmentation with small training sets applied to aerial agricultural object detection. Outcome: data insufficient. Reason: Training set size below 500 images caused accuracy to drop and fluctuate unpredictably
Application of unmanned aerial systems and deep learning in high-throughput plant phenotyping · Texas Tech · PhD
Considered and rejected
Considered and rejected: Rejected relying purely on automated deep-learning semantic segmentation for 15,988 canola images due to the prohibitive human cost of generating manual ground truth masks
Using procedural models to improve image-based plant phenotyping · Saskatchewan · PhD
Considered and rejected
Considered and rejected: Rejected semantic segmentation for patch extraction due to extreme labeling labor requirements compared to SIFT/PHOW clustering.
Unravelling the Spatial Distribution of Individual-Level Abandoned Houses at Large Scale Using Open-Access Remotely Sensed Data · University at Buffalo · PhD
Left open by the authors
Problems the authors named and did not get to.
Left open
Develop automated registration, AI feature detection, and semantic HBIM modeling algorithms for architectural documentation point clouds. Blocker: The thesis provides only broad directions without technical specifications, datasets, or concrete algorithmic architectures
Optimizing The Digital Documentation Process by Integrating Laser Scanning, Hand Measurements, And Photographic Data · Texas Tech · Master's
Left open
Evaluate Dynamic Focal Loss on object detection and semantic segmentation benchmarks using standard computer vision datasets. Blocker: None
Advancing Industrial Monitoring with Deep Learning: From Bottleneck Analysis to Safety Compliance · University of Windsor · PhD
Left open
Apply the thesis's quality quantification framework to transformer-based semantic segmentation models such as ViT and Swin Transformer. Blocker: None
Incorporating fuzzy-based methods to deep learning models for semantic segmentation · Nottingham · PhD
Left open
Weight unit-semantic pairs by an IoU threshold greater than 0.25 within GAN dissection unit count comparisons. Blocker: None
Visualizing Semiotics with Generative Adversarial Networks · Harvard · PhD
Left open
Implement semantic segmentation-based network architectures to enable pixel-wise anomaly localization for unmanned aerial systems imagery. Blocker: None
Adversarial Learning based framework for Anomaly Detection in the context of Unmanned Aerial Systems · Virginia Tech · Master's
Left open
Extend weakly supervised satellite semantic segmentation to self-supervised frameworks with automatic error detection for refining low-resolution labels. Blocker: None
Weak-Supervised Deep Learning Methods for the Analysis of Multi-Source Satellite Remote Sensing Images · University of Trento · PhD
Left open
Implement semantic segmentation to extract complex iceberg morphologies from coastal time-lapse photography instead of using bounding rectangles. Blocker: Access to the specific coastal time-lapse photography datasets and ground-truth segmentation masks.
Iceberg Stability Investigations Using Machine Learning for Alaska and Greenland · Virginia Tech · Master's
Left open
Disentangle semantic learning from geographic memorization and spatial overfitting caused by dominant prototype words in multimodal land cover mapping models. Blocker: No specific approach, metrics, or evaluation protocol are specified for disentangling spatial memorization.
Multimodal Land Cover Mapping from Remote Sensing Imagery, Species Observations, and Language · EPFL · PhD
Left open
Evaluate vision transformer architectures for off-road terrain semantic segmentation using the proposed FCIoUv2 loss function. Blocker: None
Class-Based Approach to Mitigating Class Imbalance Issues in Vision-Based Off-Road Terrain Type Recognition · Carleton University · Master's
Left open
Weight navigation graph candidate paths using long-range 2D semantic segmentation masks to penalize obstacle-dense regions. Blocker: None
Perception and Planning for Autonomous Navigation in Unstructured Environments · Cornell · PhD
Who has worked on this
Authors of the public theses cited on this page, most-cited first.
- Alex Matthew Hamer · Cranfield · PhD · 2021
Mapping agricultural land in support of opium monitoring in Afghanistan with Convolutional Neural Networks (CNNs). - Mohammed Baharoon · Harvard · Master's · 2026
AI Systems for Understanding and Grounding Radiology Reports - Luis Felipe Wolf Batista · Georgia Tech · PhD · 2026
Autonomous System for Identifying and Capturing Floating Waste - Hana Nicole Herndon · Georgia Tech · PhD · 2026
Enhancing Structural Inspections by Integrating Computer Vision, Machine Learning, and Risk Assessment for Comprehensive Corrosion Evaluation - Roozbeh Ayazi · Texas Tech · Master's · 2026
Optimizing The Digital Documentation Process by Integrating Laser Scanning, Hand Measurements, And Photographic Data - Sachiko Kaikilanialiiopuna Sakai · Virginia Tech · Master's · 2026
Iceberg Stability Investigations Using Machine Learning for Alaska and Greenland - Valérie Zermatten · EPFL · PhD · 2026
Multimodal Land Cover Mapping from Remote Sensing Imagery, Species Observations, and Language - Liao Wentong · Leibniz Universität Hannover · PhD · 2025
Scene understanding via scene graph Szenenverständnis mittels Szenengraph - H He · Oxford · PhD · 2025
Application of machine learning methods to the analysis of x-ray angiography images - Philip McEntee · University College Dublin · Doctorate · 2025
Fluorescence Imaging and Signal Quantification for Enhanced Surgical Precision and Real-Time Decision Making - Thiên-Anh Nguyen · EPFL · PhD · 2025
Monitoring and understanding treeline dynamics in the Swiss Alps from 80 years of aerial imagery - Nazifa Azam Khan · Saskatchewan · PhD · 2025
Using procedural models to improve image-based plant phenotyping
Checking a claim in this area?
We can run the same search on any method or claim. If nothing turns up, we will say so, and that proves nothing on its own.