M.Sc. Data Science. Machine learning for Earth observation and remote sensing. Berlin, Germany. Looking for a PhD position or research role in Earth observation, remote sensing or GeoAI. Available immediately.

_{Left: SAR calving front from a frozen photo-pretrained encoder (CaFFe, Sentinel-1). Middle: VGGT-Ω depth error on a driving scene, black rings mark moving objects (FZI-AURA). Right: new buildings found from frozen DINOv3 features (LEVIR-CD).}

	sar-transfer-glaciers — Do frozen encoders pretrained on optical images transfer to SAR? Glacier calving-front delineation on the CaFFe benchmark with frozen DINOv3 (satellite and photo), C-RADIOv4-H and two SAR-pretrained encoders, only small probes and heads trained on top, a U-Net from scratch as reference. A 0.57 M-parameter decoder on frozen C-RADIO features matches the 7.8 M-parameter U-Net on front error on the test glaciers (979 ± 31 m vs 979 m). Photo-pretrained encoders beat the SAR-pretrained ones we tried, and satellite pretraining gave no clear advantage over photo pretraining.
	vggt-omega-aura-benchmark — Failure analysis of the released VGGT-Ω 3D reconstruction checkpoint on FZI-AURA, a driving dataset published after the model and so outside its training data. Depth and pose error are broken down by weather, lighting and object motion instead of averaged. Daytime depth on the held-out test split reaches AbsRel 0.082; error concentrates on thin objects, people and moving vehicles. Python reference implementation with a C++ core for the scoring.
	Change-Detection-Using-Dinov3 — Building change detection on frozen satellite-pretrained DINOv3 features, on SpaceNet-7 monthly imagery and LEVIR-CD. Decoder designs compared on identical features: feature differencing at 0.56M parameters and cross-attention at 0.83M finished within 0.002 F1 of each other at 0.910 on LEVIR-CD, so the smaller head was the one to keep. Both beat the frozen embeddings read directly by roughly ten times on SpaceNet-7. Trained on a free Colab T4.
	owl-caribou-overhead — Cross-herd, cross-year evaluation of four overhead animal detectors on 2,607 aerial survey patches of the Central Arctic caribou herd (Alaska, 2022), with error bars from resampled mosaics, a failure analysis and a threshold sweep. The two best models tie threshold-free (average precision 0.978 vs 0.977); the gap at the fixed setting is an operating-point effect. Half of the remaining errors sit in a 16 px band at the patch border, and most of those are real animals the patch's ground truth does not list.
	bavaria-wheat-sentinel2-oco2 — Monthly NDVI and NIRv for winter wheat across the 96 NUTS-3 regions of Bavaria, 2017–2024, from 1.13 TB of Sentinel-2 L3A composites masked with yearly 10 m crop type maps, aggregated with exact partial-pixel zonal statistics on the GPU. The Sentinel-2 half of a two-person MSc capstone. A side analysis matches Sentinel-2 reflectance to OCO-2 solar-induced fluorescence footprints.

Carried out at the DLR Earth Observation Center, Oberpfaffenhofen: detection of greenhouses and plastic-covered parcels in southern Germany from 20 cm aerial orthophotos, with no hand-drawn annotation anywhere in training. Labels were derived from EU parcel-level crop declarations and refined against the imagery. The detector combines a frozen satellite-pretrained DINOv3 backbone, decoded back to 20 cm by guided feature upsampling, with a trainable ResNet-34 branch and a small fusion head. Supervised by Dr Ursula Gessner (DLR) and Prof Dr Iftikhar Ahmed (UE). A paper is in preparation.

Most of my projects derive training labels from registers or benchmarks that were never made for the purpose, then deal with the biases those labels carry. A result that has come back in every project so far: past a small model budget, improving the data moves accuracy further than adding model capacity does, and I measure both sides rather than assume it. Pipelines are numbered stages with pinned environments so someone else can rerun them, and evaluation is on sites and conditions held out from training, not on random splits.

• Student Assistant, Robotics and Perception, TU Berlin, 2025–2026. Onboard image capture and control components in Python for REINCARNATE, a Horizon Europe consortium project.
• Test Engineer, Infosys, India, 2021–2024. Automated test pipelines in Java and JavaScript with Selenium and TestNG, and data validation in SQL.

• M.Sc. Data Science, University of Europe for Applied Sciences, Potsdam, 2024–2026.
• B.Eng. Electronics and Communication Engineering, Jaypee Institute of Information Technology, India, 2016–2020.

Python, PyTorch, C++17 (CMake, pybind11), SQL. GDAL, rasterio, geopandas, xarray, QGIS, Google Earth Engine, STAC. Linux, Git, Docker, pinned environments. GPU training on A100 and HPC on LRZ terrabyte (SLURM).

• LinkedIn: abhishekzsingh
• Website: saverin0.github.io

_{Figure data: CaFFe (Gourmelon et al. 2022, CC BY 4.0), FZI-AURA, LEVIR-CD, OWL caribou survey release (CC BY-NC-SA 4.0), DLR Sentinel-2 composites.}