How hehaoyang1124/aimer works
AImer provides two Windows aim-assistance applications: a configurable deep-learning application for supported games, and an Aimlabs Gridshot assistant using traditional computer vision. The principal workflow selects a game and target, captures screen frames, detects targets, computes mouse corrections, and sends mouse input; Aimlabs follows a separate capture-and-detect loop. The README documents ONNX Runtime/CUDA and OpenVINO inference options and configurable game/model support. Sampled source confirms screen capture, launcher entry, and mouse-driver selection; detector and launcher internals were not sampled, so their detailed wiring is not inferred.
Main entry points: Launcher.cpp Launcher UI · autoAim.cpp DL aim loop · ScreenCapture.cpp DXGI capture · Detector.cpp Detector interface
Components 17
DL application
- DL entry
src/AImer-DL/main.cpp - Launcher UI
src/AImer-DL/launcher/Launcher.cpp - Game settings
src/AImer-DL/Games.cpp - DL aim loop
src/AImer-DL/autoAim.cpp
Capture and detection
- DXGI capture
src/capture/ScreenCapture.cpp - Detector interface
src/AImer-DL/detector/Detector.cpp - ONNX inference
src/AImer-DL/detector/OnnxDetector.cpp - OpenVINO inference
src/AImer-DL/detector/OpenvinoDetector.cpp
Aimlabs application
- Aimlabs entry
src/AImer-aimlab/main.cpp - Sphere detection
src/AImer-aimlab/autoAim.cpp - Aimlabs correction
src/AImer-aimlab/autoAim.cpp
Mouse input
- Mouse controller
src/controller/MouseController.cpp
Other components
- User
- Game screen
- Model files
- Input simulator
deps/IbInputSimulator/InputSimulator.hpp - Game mouse input
Connections 20
- User → DL entry: starts
- DL entry → Launcher UI: initializes
- User → Launcher UI: selects options
- Launcher UI → Game settings: loads settings
- Launcher UI → DL aim loop: launches
- Game screen → DXGI capture: provides frames
- DXGI capture → DL aim loop: supplies frames
- DL aim loop → Detector interface: requests detection
- Detector interface → ONNX inference: uses backend
- Detector interface → OpenVINO inference: uses backend
- Model files → ONNX inference: loads model
- Model files → OpenVINO inference: loads model
- User → Aimlabs entry: starts
- Aimlabs entry → DXGI capture: captures via
- DXGI capture → Sphere detection: supplies frames
- Sphere detection → Aimlabs correction: returns offset
- Aimlabs correction → Mouse controller: requests movement
- DL aim loop → Mouse controller: requests movement
- Mouse controller → Input simulator: sends input
- Input simulator → Game mouse input: simulates